2014 gas__

gas 2014

PTQ supplement

gas cov copy 6.indd 1 06/03/2014 13:57

1215

_e

www.engineering-solutions.airliquide.com

Because our genes lead us to an irrepressible passion for new solutions.

The only thing longer than our company historyis the list of our innovations.

When it comes to providing innovative plant and system solutions, Air Liquide Global E&C Solutions is your partner of choice – worldwide. No surprise, since we innovate relentlessly in order to bring sustainable and cost-eff ective solutions

to society, levera ging partnerships with customers, suppliers, academics and communities. This is not only demonstrated by thousands of patents, but by pioneering reference projects around the globe as well. Yours could be the next one.

AIR2362022_AZ_Techno-DNA_7-8-2013_RZ.indd 1 21.05.13 11:26gas air liquide.indd 1 04/03/2014 10:44

http://www.eptq.com/jump.aspx?a=Air Liquide&p=gas14

©2014. The entire content of this publication is protected by copyright full details of which are available from the publishers. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means – electronic, mechanical, photocopying, recording or otherwise – without the prior permission of the copyright owner.The opinions and views expressed by the authors in this publication are not necessarily those of the editor or publisher and while every care has been taken in the preparation of all material included in Petroleum Technology Quarterly and its supplements the publisher cannot be held responsible for any statements, opinions or views or for any inaccuracies.

3 China’s sour gas ChrisCunningham

5 Zero emissions in sulphur recovery MicheleColozziandSimonaCortese KT-Kinetics Technology LuciaBarbato Processi Innovati 15 Eliminating the Claus furnace MattThundyil,SameerPallavkar,RamiroVazquezandDavidSeeger GTC Technology US, LLC

31 Zeolite based dryers in ethylene plants VassiliosZafirakisandHansHoefer Grace Materials Technologies 49 Virtual commissioning of a gas handling system RainerScheuring Cologne University of Applied Sciences Hans-ChristianHaarmann-KühnandJürgenEssler TGE Marine Gas Engineering TorstenFelix Flemming Automationstechnik MichaelBrodkorb Honeywell Process Solutions 54 Protecting compressors with dynamic simulation NicholasBrownrigg AspenTech

WoodsidePetroleum’sPlutoLNGprojectinWesternAustralia.Photo: Woodside Petroleum

2014www.eptq.com

gasptqYLRETRAUQYGOLONHCET MUELORTEP

ed com copy.indd 1 12/03/2014 12:43

ProMax is a powerful and versatile process simulation package that is used worldwide to design and optimize

integrated with Microsoft Visio®, Excel® and Word®,ProMax is the best simulation tool for designing gas

Discover why engineers around the world depend on ProMax!

2

2

Process Focus: Gas Processing

ProMax® with TSWEET® & PROSIM®

Process Simulation Software

bre.indd 1 8/3/12 10:09:31

http://www.digitalrefining.com/61,sponsors,Bryan_Research___Engineering.html

The European Union has arguably been the global leader in biodiesel production and use, with overall

biodiesel production increasing from 1.9 million tonnes in 2004 to nearly 10.3 million tonnes in 2007. Biodiesel production in the US has also increased dramatically in the past few years from 2 million gallons in 2000 to approximately 450 million gallons in 2007. According to the National Biodiesel Board, 171 companies own biodiesel manufacturing plants and are actively marketing biodiesel.1. The global biodiesel market is estimated to reach 37 billion gallons by 2016, with an average annual growth rate of 42%. Europe will continue to be the major biodiesel market for the next decade, followed closely by the US market.

Although high energy prices, increasing global demand, drought and other factors are the primary driv-ers for higher food prices, food competitive feedstocks have long been and will continue to be a major concern for the development of biofu-els. To compete, the industry has responded by developing methods to increase process efficiency, utilise or upgrade by-products and operate with lower quality lipids as feedstocks.

Feedstocks

Biodiesel refers to a diesel-equivalent fuel consisting of short-chain alkyl (methyl or ethyl) esters, made by the transesterification of triglycerides, commonly known as vegetable oils or animal fats. The most common form uses methanol, the cheapest alcohol available, to produce methyl esters. The molecules in biodiesel are primar-ily fatty acid methyl esters (FAME), usually created by transesterification between fats and methanol. Currently, biodiesel is produced from various vegetable and plant oils. First-genera-tion food-based feedstocks are straight vegetable oils such as soybean oil and animal fats such as tallow, lard, yellow grease, chicken fat and the by-products of the production of Omega-3 fatty acids from fish oil. Soybean oil and rapeseeds oil are the common source for biodiesel production in the US and Europe in quantities that can produce enough biodiesel to be used in a commercial market with currently applicable technologies.

First-generation feedstocks for

Gas 2014 3

Editor Chris Cunningham [email protected]

Production EditorRachel [email protected]

Graphics EditorRob [email protected]

Editorial tel +44 844 5888 773fax +44 844 5888 667

Business Development DirectorPaul [email protected] Advertising SalesBob [email protected]

Advertising Sales Officetel +44 844 5888 771 fax +44 844 5888 662

PublisherNic [email protected]

CirculationJacki [email protected]

Crambeth Allen Publishing LtdHopesay, Craven Arms SY7 8HD, UKtel +44 844 5888 776fax +44 844 5888 667

ISSN 1362-363X

ptq (petroleum technology quarterly) (ISSN No: 1632-363X, USPS No: 014-781) is published quarterly plus annual Catalysis edition by Crambeth Allen Publishing Ltd and is distributed in the US by SP/Asendia, 17B South Middlesex Avenue, Monroe NJ 08831. Periodicals postage paid at New Brunswick, NJ. Postmaster: send address changes to ptq (petroleum technology quarterly), 17B South Middlesex Avenue, Monroe NJ 08831.Back numbers available from the Publisher at $30 per copy inc postage.

China’s sour gas

China has set about its plan to step up consumption of gas purposefully, in a bid to cut levels of air pollution. For the time being, that means accelerating imports as well as raising gas prices to encourage domestic

development.The PRC’s gas imports are sourced as liquefied natural gas, chiefly from

Africa, and pipeline gas from central Asia. At the beginning of March, the China Development Bank approved a $700 million loan to complete a section of the Kazakhstan-China main gas pipeline, itself part of a vast network of gas-carrying pipelines linking the southern former Soviet republics and western China.

At the same time, domestic energy majors Sinopec and CNPC said that they were likely to meet a state annual target of 6.5 billion m3 of shale gas in 2015, rising to as much as 100 billion m3 by the end of the decade. These figures compare with total production in 2013 of 200 million m3, so there is some reliance on unproven reserves of shale gas.

China also has vast conventional reserves of natural gas, in particular in the south western province of Sichuan and in Xinjiang to the west. So why is domestic gas not already piped into the nation’s industrial centres rather than from central Asia or via LNG terminals? For one thing, China’s onshore gas reserves are classed as ‘very sour’ with a hydrogen sulphide content of, typically, 15%. That is a high but not prohibitive barrier to gas processing. Crucially, these are also high pressure fields. An accidental release of untreated gas has the potential for disaster, as proved to be the case over a decade ago when a blow-out poisoned the city of Chongqing.

Nonetheless, Sinopec has operated a safe production and processing site in the region since 2010. The next big development is the Chuandongbei sour gas project, run by Chevron and CNPC under a production sharing contract. The project was originally due onstream in 2010, but has been beset by the technical difficulties of recovering high-pressure sour gas and severe cost over-runs. Chuandongbei is now expected to start up in 2015 and is slated to reach daily production of over 500 000 cubic feet of pipeline quality gas.

Chevron is the only international oil major operating in China’s gas fields, in conventional onshore fields, in shale deposits, and offshore. The compa-ny’s total production of gas in China is reported at 20 000 b/d oil equivalent in 2013. There is far to go for China’s natural gas industry.

Should China move ahead more purposefully with its sour gas processing, there is a potential down-side for refinery costs worldwide. Throughout the current century, China has been the world’s largest importer of sulphur. The PRC’s fertilizer companies convert this to sulphuric acid for the manufacture of ammonium phosphate which has transformed the nation’s level of self- reliance in the production of crops.

Most of this sulphur for China originates in Canada’s sour gas processing industry, as well as the Middle East’s oil refineries and gas production sites, Kazakhstan’s Caspian developments and US West Coast refineries. It is a market that provides a welcome outlet for — from the refiners’ perspective — a largely unwanted, low-value byproduct. The eventual disappearance of a China market for sulphur will add to some refiners’ disposal problems.

CHRIS CUNNINGHAM

ptqYLRETRAUQYGOLONHCET MUELORTEP

ed com copy.indd 2 06/03/2014 12:16

Visit our webpage to learn more about all our products > www.jacobs.com/comprimo

Comprimo® Sulfur Solutions

Total Project SolutionsWe are one of the world’s largest and most diverse providers of professional technical services. Our global network includes more than 200 offices in more than 25 countriesFrom our gas treating & sulfur technology centers of expertise in Leiden, The Netherlands and Calgary, Canada, we provide engineering, procurement and construction services to our clients around the world

Gas Treating & Sulfur Technology

With more than 500 units built during the last 40 years, we are a global leader in the gas treating and sulfur recovery technology

gas jacobs.indd 1 04/03/2014 10:46

http://www.digitalrefining.com/173,sponsors,Jacobs_Comprimo_Sulfur_Solutions.html

Zero emissions in sulphur recovery

Environmental regulations around the world nowadays impose very strict SO2 emis-

sions requirements. H2S flaring is no longer tolerated in industrial complexes; best available technique (BAT) has to be utilised to satisfy such new requirements. The achievement of a high level of relia-bility and availability is also mandatory for sulphur recovery units (SRU) installed in fully inte-grated industrial complexes where production must be maximised.

The main targets of the SRU to be installed in a new industrial complex can be summarised as follows:• SO2 emissions no higher than 150 mg/Nm3 (50 ppmv)• Ensure complete destruction of impurities contained in sour gas

A process under development for sulphur recovery features an innovative catalyst and a new scheme for the treatment of all sour gas feedstocks

MICHELE COLOZZI and SIMONA CORTESE KT-Kinetics Technology LUCIA BARBATO Processi Innovati

feedstocks (NH3, HC, COS, CS2, RSH, HCN, and so on)• Maximum energy recovery• Maximum availability and reliability• Minimum opex and capex.

The Modified Claus section can achieve a sulphur recovery effi-ciency (SRE) of 92-96% when provided with two reactors, and 94-98% when provided with three reactors. Therefore, the addition of a tail gas treatment (TGT) section is always required to achieve a SRE higher than 99.9%.

In the 1990s, Kinetics Technology SpA (KT) introduced the proprie-tary RAR (reduction, absorption, recycle) technology which is recog-nised as BAT and is specifically developed to minimise SO2 emis-sions with the aim of being in full

compliance with the most stringent environmental regulations in force. The RAR process is a reductive amine based tail gas treatment process, capable of achieving the maximum mandated sulphur recovery efficiency (SRE>99.9+%). A block diagram of the RAR process is shown in Figure 1.

In the proprietary KT tail gas treatment section, the Claus tail gas is mixed with hydrogen rich gas from battery limits, when available, or it can be produced in a dedi-cated reducing gas generator (in-line heater). Tail gas is then sent to the hydrogenation reactor, where all components containing sulphur are reduced and/or are hydrolysed to H2S.

Tail gas is then sent to a quench tower where, through direct contact

www.eptq.com Gas 2014 5

Sour water stripper acid gas

Hydrogen make-up

Amine acid gas

Reduction

Absorption RecycleModified Claus

section

Liquid sulphur degassing

section

Heating/cooling section

Amine regeneration

section

Incineration section

Fuel gasFlue gas

Sour water

Liquid sulphur

Claus tail gas

Lean amine

Rich amine

Acid gas recycle

Super enrichment

Off gas recycle

TGT tail gas

Sweet tail gas

Vent gas recycle

Undegassed liquid sulphur

Figure 1 Block diagram of the RAR process

gas kt.indd 1 06/03/2014 12:21

6 Gas 2014 www.eptq.com

RAR process is shown in Figure 2.

Integrated sour gas selectiveoxidative autothermal processEnvironmental regulations will become more and more stringent in the near future, requiring SO2 emis-sions to the atmosphere to be much lower than 150 mg/Nm3 (lower than 50 ppmv). Therefore, new process schemes will be required in order to be in compliance with the strictest future regulations. For this

with circulating water, steam is condensed and the gas is cooled down. The H2S contained in the tail gas is removed by the use of an amine solution in the TGT absorber. Sweet tail gas from the top of the TGT absorber, containing a few parts per million of H2S, is sent to the thermal incinerator to be converted to SO2 before the flue gases are released to the atmos-phere. The amine solution is normally regenerated in a dedicated

amine regeneration section, included in the sulphur recovery unit. The produced liquid sulphur is treated in the liquid sulphur degassing section where hydrogen polysulphides and H2S are completely removed. In order to minimise SO2 emissions to the atmosphere, all of the gases released during liquid sulphur degassing are recycled back to the Claus section and the TGT section. A typical process flow scheme of the

SWS acid gas

Acid gas

CA

Thermal reactor

Sulphur condensers

Claus stage I

RAR

WHB

Thermal incinerator

WHB

Stack

Claus stage II

Sulphur

Rich amine

Flue gas

H2CA

Fuel gas

Acid gas recycle

Figure 2 Typical process flow scheme of the RAR process

Sour water stripper acid gas

O2

Reduction

Absorption RecycleH2S cracking section

(SOAP section)

Liquid sulphur degassing

section

Heating/cooling section

Amine regeneration

section

Incineration section

(Start-up only)

Fuel gasFlue gas to ATM

Hydrogen rich gas

Sour water

Liquid sulphur

Claus tail gas

NNF (for start-up only)

Lean amine

Rich amine

Acid gas recycle

Super enrichment

Off gas recycle

TGT tail gas

Vent gas recycle

Undegassed liquid sulphur

Amine acid gas

Figure 3 Block diagram of the new concept SRU

gas kt.indd 2 06/03/2014 12:21

Gas Processing PlantWestern USA

A World of Solutions Visit www.CBI.com

SHALE GAS PROCESSING SOLUTIONSComplete Capabilities Addressing the Entire Natural Gas Value ChainFor more than 40 years, CB&I has excelled at creating innovative solutions for the gas processing industry, employing a deep portfolio of execution capabilities and processing technologies. Drawing upon our full-service technology plus EPFC expertise, CB&I has installed more than 600 gas processing related projects.

From concept through commissioning across the entire natural gas value chain, we deliver. Contact CB&I to learn more about our modular gas processing plants and superior gas technologies and how we can improve your next gas processing project.

FIELD DEVELOPMENTGAS CONDITIONING AND TREATINGSULFUR REMOVAL AND RECOVERYFRACTIONATIONNITROGEN REJECTIONDEEP ETHANE AND LPG RECOVERYLIQUID PRODUCT TREATINGUPGRADING OF LOW VALUE GASMODULARIZATION

cbi_ptq_gas_suppl_ad_apr_2014.indd 1 3/6/2014 1:37:20 PM

cbi.indd 1 07/03/2014 09:51

http://www.digitalrefining.com/64,sponsors,CB_I_.html

reason, KT is developing an inno-vative process and a new concept for the SRU.

The process scheme under devel-opment is divided into two main sections:• H2S cracking section (sour gas SOAP section)• Tail gas treatment section (amine based).

The core part of the innovative SRU is the Sour Gas Selective & Oxidative Autothermal Process (Sour Gas SOAP) section which is fully integrated with the RAR tail gas treatment plant’s configuration and process.

A block diagram of the novel process is shown in Figure 3.

The new concept SRU developed by KT is a simplified scheme equipped only with a catalyst-filled reaction furnace package (catalytic reaction furnace or CRF) followed by one sulphur condenser, a liquid sulphur degassing section and a TGT section.

What is the advantage of sucha configuration?The reason for such a minimal configuration is strictly related to the novel idea of operating the Claus section in a different way:• Use of a proprietary catalyst in the reaction furnace to minimise SO2 production


• Use of a specific H2S/O2 ratio in order to maximise production of hydrogen while minimising the production of SO2• Use of oxygen instead of ambient air.

The peculiarity of the process is the possibility to install a catalyst block directly in the reaction furnace, in order to enhance the selective conversion of H2S directly to H2, Sx and H2O instead of Sx, SO2 and H2O.

An interesting advantage of the new concept SRU is the possi-bility to achieve important capex reductions, minimising the number of items of equipment to be installed in the plant (see Figure 4).

Additionally opex can be reduced with the process since a thermal incinerator is virtually no longer required in normal operations, the hydrogen rich gas produced can be utilised in a petroleum refinery, and SO2 and CO2 emissions are drastically reduced. In natural gas production, where hydrogen users are not available, the operating parameters may be adjusted to minimise production of hydrogen rich gas.

Reaction mechanismH2S cracking through partial oxida-tion in the CRF using a proprietary

catalyst is based on the following reactions:

H2S + 3/2 O

2 → H

2O + SO

2

(oxidation reaction) (1)

H2S + 1/2 O

2 → 1/2 S

2 + H

2O

(partial oxidation reaction) (2) 2H

2S + SO

2 ↔ 2H

2O + 3/n Sn

(Claus reaction) (3)

H2S ↔ H

2 + 1/2 S

2 (cracking reaction) (4)

KT is developing the proprietary catalyst in collaboration with the University of Salerno (UNISA) in Italy. The laboratory plant to test the proprietary catalyst and the novel process is located at the University of Salerno. KT has selected UNISA because it has the best facilities in Europe on an academic level as well as the high-est competence in the field of sour gas heterogeneous catalysis. More than 4000 hours of tests in the labo-ratory plant at UNISA have been performed. For feedback of the results of laboratory tests, KT has developed a thermodynamic and kinetic model which fits with high accuracy the results of the labora-tory plant. The model utilises more than 100 reactions and in this way it approaches actual system behaviour.

SWS acid gas

Acid gas

CA

Catalyticreactorfurnace

Sulphur condensers

Claus stage I

RARH2 rich gas

WHB

Thermal incinerator

WHB

Stack

Claus stage II

Sulphur

Acid gas recycle

Rich amine

Flue gas

H2CA

Fuel gas

Figure 4 Process flow scheme for the new concept SRU

gas kt.indd 3 11/03/2014 17:01


H2; CO and H2S are the fi nal prod-ucts of reaction.

KT has evidence of such behav-iour in all of the laboratory tests performed.

New concept SRU for industrial applicationsBased on knowledge acquired from the laboratory tests, KT has

and CS2 in the gas phase at the outlet of the catalytic reaction furnace. The COS and CS2 destruc-tion mechanism is based on the adsorption of hydrocarbons and H2S over the catalyst surface. The hydrocarbons and H2S then react over the catalyst producing CS2. The CS2 produced reacts with SO2 producing COS which reacts with

Figures 5 to 7 show the main mechanisms of reaction involved in the SOAP Process in the presence of the catalyst.

H2S reaction mechanism

The heat for cracking reaction (4) is made available by the oxidation reaction (1 & 2) of H2S. The oxygen provided to the system is only suffi -cient to achieve the proper reaction temperature to activate all the rele-vant reactions over the catalyst.

A part of the H2S is oxidised to SO2and H2O (1) and another part of H2S is partially oxidised to Sx and H2O (2) producing the heat required for the other reactions. At the same time, part of the H2S adsorbed over the catalyst reacts with SO2 according to the Claus reaction (3), then the other part of the H2S produces H2 by cracking reaction (4).

KT has evidence from laboratory tests that, by utilising its proprie-tary catalyst, the cracking reaction (4), the partial oxidation reaction (2) and the Claus reaction (3) are favoured, while SO2 formation is negligible.

NH3 reaction mechanism

One of the major merits of the novel process is the possibility to treat a very high content of ammo-nia even without the presence of H2S. NH3 destruction follows the same mechanism as H2S conver-sion, therefore this reaction is also conducted autothermally. A part of the ammonia is oxidised with the oxygen provided to the system to produce N2 and H2O, in turn supplying the heat required for the NH3 cracking reaction producing N2 and H2.

This behaviour is very interesting for the future treatment of extra heavy crudes that will be increas-ingly utilised in the coming years.

Hydrocarbons, COS and CS2

reaction mechanismA peculiarity of the catalyst is that hydrocarbons adsorbed over the catalyst will react preferably and exclusively with H2S, therefore no coke formation is expected. In this way it is possible to avoid completely the presence of COS

KT catalyst

O2 (g)

Cracking

H2S (g)

H2S (ads) H2S (ads)H2 (g) + S2 (g)

SO2 (g) + H2O (g)

S2 (g) + H2O (g)

S2 (g) + H2O (g)

Partial oxidation

Claus r

eact

ion

Tota

l oxi

datio

n

Figure 5 H2S reaction mechanism over proprietary catalyst

KT catalyst

O2 (g)

Cracking

NH3 (g)

NH3 (ads)H2 (g) + N2 (g)

N2 (g) + H2O (g)

Tota

l oxi

datio

n

Figure 6 NH3 reaction mechanism over proprietary catalyst

KT catalyst

O2 (g)

Total oxidation

Partial oxidation

Claus r

eact

ion

Par

tial o

xida

tion

CrackingO2 (g)

H2 (g)

S2 (g)CH4 (g)

CH4 (ads)

CH4 (g)

H2S (g)

H2S (ads)H2S (ads)

CS2 (ads)

COS (ads)

H2 (g) + S2 (g)

SO2 (g) + H2O (g) S2 (g) + H2O (g)

S2 (g) + H2O (g)

CO2 (g) + H2O (g)

CO(g) + H2S (g)

CO(g) + H2 (g)

H2S(ads) + CH4 (ads)

Tota

l oxi

datio

n

Figure 7 Hydrocarbons, COS and CS2 reaction mechanism over proprietary catalyst

gas kt.indd 4 06/03/2014 12:21

This is not a checkerwall. This is not a choke ring.

This is a VectorWallTM and it will change the way you look at your reaction furnace.

With unmatched mechanical stability, the VectorWallTM will increase your furnace capacity and eliminate ammonia carryover, saving your company millions of dollars.

For more information check out our new video at petrochemical.blaschceramics.com Or contact us: 1.518.436.1263 [email protected]

Our VectorWallTM is saving SRU reaction furnace operators

millions of dollars

Jeffrey J. Bolebruch Senior Market Manager

blasch.indd 1 11/03/2014 16:31

http://www.digitalrefining.com/161,sponsors,Blasch_Precision_Ceramics.html


New concept SRU in gas productionThe results of opex analysis for a gas field application comparing conventional and new concept SRUs are shown in Table 2 and Figure 9.

According to the selected plant configuration, capex reduction of up

process and in the TGT section. Low pressure steam for amine regeneration is imported from the battery limit while the excess steam is sent to a high pressure and/or medium pressure steam header according to relevant network conditions.

performed process simulations of the novel process with high accu-racy, with the purpose of choosing the best process scheme and the best operating conditions. In this regard it was possible to perform several case studies for different industrial applications including:• Petroleum refining• Oil and gas production• Coal gasification• Integrated gasification combined cycle (IGCC) complex• Chemical and petrochemical complex.

In this article, we present the results of our opex and capex anal-ysis relevant to a petroleum refinery and a gas field.

New concept SRU in a petroleum refineryThe results of opex analysis for a petroleum refinery application, comparing conventional and new concept SRUs, are shown in Table 1 and Figure 8.

A conventional plant configura-tion consists of a Modified Claus section equipped with two Claus reactors followed by an amine-based reductive TGT section with a dedi-cated amine regeneration system and the relevant thermal incinera-tion section. A liquid sulphur degassing section is also foreseen.

The novel process is equipped only with a CRF system and a sulphur condenser followed by an amine-based reductive TGT section with a dedicated amine regenera-tion system, and a liquid sulphur degassing section with full recycle of H2S to the Claus and TGT sections. A thermal incinerator is also foreseen but it is to be utilised only during start-up and shutdown operations (see Figure 4).

In plant configuration A, part of the high pressure steam produced is utilised as the heating medium in the novel process and in the TGT section, while the excess steam is let down to low pressure steam which is utilised as the heating medium for amine regeneration. With this proposed configuration, steam export will not be provided.

Plant configuration B utilises the produced high pressure steam as the heating medium in the novel

40

8070

10090

6050

302010

0−10

Pro

port

ion

, %

−20

LPS

MPS

Hydro

gen ri

ch g

as

CO 2 e

miss

ions

Fuel g

as

Elect

ric p

ower

Conventional SRU conceptPlant config. A of new SRU conceptPlant config. B of new SRU concept

Figure 8 Opex analysis for a SRU installed in a petroleum refinery

SRU conventional Plant configuration A Plant configuration B concept new SRU concept new SRU conceptLow pressure steam import Yes Yes YesMedium pressure steam export Yes No YesHydrogen rich gas No Yes YesCO

2 emissions Yes No No

Fuel gas Yes Yes YesElectrical power Yes Yes YesProfit index, %* 26 34 100

*Profit = (total cost of utilities)/(ton of sulphur produced) Profit index = (100 x profit) / (profit of plant configuration with maximum. profit)

Opex analysis for a SRU installed in a petroleum refinery

Table 1

SRU conventional concept New SRU concept Low pressure steam import Yes YesMedium pressure steam export Yes NoHydrogen rich gas No NoCO

2 emissions Yes Yes

Fuel gas Yes YesElectrical power Yes YesProfit index, %* -100 -74

*Profit = (Total cost of utilities)/(ton of sulphur produced) Profit index = (100 x profit)/(profit of plant configuration with maximum profit)

Opex analysis for a SRU installed in a gas field

Table 2

gas kt.indd 5 06/03/2014 12:21


TGT is 99.9++%, much higher than the 99.9+% of current BAT. The reason for this difference is the fact that in the novel process the quan-tity of total sulphur (H2 S+COS+CS2) in the top of the TGT absorber is much lower than in the conventional one•The same flow scheme for theSRU can be used in all industrial applications.

Up to now only a few disadvan-tageshavebeenidentified:• The H2 S cracking section has a lower sulphur recovery efficiencywhen compared with the conven-tional one which has a reaction furnace and two or three additional Claus reactors• In case of shutdown of the TGT section, a higher quantity of SO2 may be released to the atmosphere. Nevertheless such a scenario is very rare or near-impossible due to the very high availability and reliability of the TGT section, especially when treating tail gas from the H2 S crack-ing section with zero SO2 content• The availability and reliability of the proprietary catalyst needs to be demonstrated in a pilot plant.

This article is based on a paper presented at the ADIPEC 2013 Technical Conference, Abu Dhabi, UAE, 10-13 November 2013.

Michele Colozzi is Business Development and Technology Manager with KT-Kinetics Technology SpA (KT) in Rome and is involved in the commercialisation, design and development of new processes, construction, commissioning and start-up of sulphur and gas treatment plants, oil and gas treatment plants, hydrogen and synthesis gas generation plants, and waste disposal plants. He holds a chemical engineering degree from University of Rome “La Sapienza”. Simona Cortese is a Business Development Coordinator with KT-Kinetics Technology. She works as coordinator of the technical team for the development of gas treatment and sulphur recovery technical proposals for PDP and FEED activities and as coordinator of research and development team. She graduated as a chemical engineer from the University of Rome “La Sapienza”. Lucia Barbato is a Process Engineer with Processi Innovativi Srl, a process company specialising in R&D and owned by KT-Kinetics Technology. She is involved in the research and development of projects for partial oxidation of hydrocarbons and H

2S, and

graduated cum laude as a chemical engineer from the University of Salerno.

ppmv) which is not possible with the current BAT• There is complete destruction of COS, CS2, hydrocarbons, RSH, ammonia, HCN and other impuri-ties in the catalytic reaction furnace• The novel process can treat only sour water stripper acid gases, without any limitation on the ammonia content in the sour gases; this is not possible with the conven-tional process• A conventional plant requires two or three Claus catalytic converter systems (pre-heater, reac-tor, sulphur condenser) to minimise the SO2, COS and CS2 content in the tail gas to the TGT section. The novel process requires only one catalytic converter system since the catalyst is installed directly in the catalytic reaction furnace. Claus catalytic converters are no longer necessary due to the low content of SO2 in the process gas from the CRF system• The novel process will produce a lower quantity of exhaust catalyst compared to the conventional process• The reduced number of items of equipment makes the investment cost of the novel process cheaper than the conventional one. In addi-tion, the associated maintenance costs will be reduced• The major advantage of the novel process is the higher overall SRE. Indeed, the SRE of H2S cracking +

to 30% may be achieved due to the reduced number of items of equip-ment foreseen in the SOAP process.The innovative plant configura-

tion shows an important peculiarity, the possibility to mini-mise SO2 emissions, since the total sulphur (H2S+COS+CS2) contained in the gas from the TGT absorber is much lower than in the conven-tional one, as explained in the followingformula:

[(STot)Top TGT Absorber

]Novel Process

<1 [ (STot)Top TGT

Absorber]Conventional Process

n

5<n<10

Utilising always the same plant configuration — the new SRUconcept (SOAP+RAR)— it will bepossible to achieve SO2 content in gas released to the atmosphere lower than 50 mg/Nm3 (18 ppmv).

ConclusionsThe sour gas SOAP is an environ-mentally friendly solution capable of valorising the sour gases which are considered a problem in all industrial complexes, producing hydrogen as valuable product, where required, with the purpose of achieving zero emissions.

The main advantages of the novel processaresummarisedbelow:• In terms of SO2 emissions, the novel process is able to achieve values lower than 50 mg/Nm3 (18

60

100

90

80

70

50

40

30

20

10

Pro

port

ion,

%

0

LPS

MPS

Hydro

gen ri

ch g

as

CO 2 e

miss

ions

Fuel g

as

Elect

ric p

ower

Conventional SRU conceptNew SRU concept

Figure 9 Opex analysis for a SRU installed in a gas field

gas kt.indd 6 11/03/2014 16:30

Copyright © 2013 Merichem Company

To win a race you need not only speed, but a timely start. Merichem Company has a proven track record of providing hydrocarbon treating solutions that meet the stringent timelines of our customers. Get your treating project off to a timely start by choosing Merichem Company as your technology licensor.

Midstream SolutionsProven Results

Think Fast...Think Merichem

Midstream Solutions with Industry-Leading Speed to Market

Merichem Company’s portfolio of midstream treating technologies

includes our patented FIBER FILM ® Contactor that offers a variety of

caustic, amine and acid treating processes. Merichem also offers

solutions for gas treating, including our patented LO-CAT ® wet scrubbing,

liquid redox system that converts H2S to innocuous, elemental sulfur. In

addition to our large portfolio of licensed treating technologies,

Merichem offers industry leading spent caustic management services that

utilize Beneficial Reuse options to create a non-waste solution for your

spent caustics. For fast, effective treating solutions, choose Merichem.

www.merichem.com | 5455 Old Spanish Trail | Houston, Texas 77023 | 713.428.5000

®

C

M

Y

CM

MY

CY

CMY

K

Merichem Company-FY14-MID.pdf 1 2/24/2014 8:19:32 AM

merichem.indd 1 06/03/2014 12:04

http://www.digitalrefining.com/91,sponsors,Merichem_Company.html

cleaner, better shale gas & oil

At Zeeco, we understand the everyday realities of shale gas & oil.

From our field-proven, smokeless enclosed or open flares to thermal

oxidisers specifically designed to handle amine/BTEX off-gasses, we

have the solutions that keep your people and your facilities safe – and

keep you in compliance. Solutions engineered for your purposes,

available from standardised inventory, and backed up by the world

leader in combustion technology. Choose the company where the

experts answer the phone and the engineers understand your needs.

Zeeco: We work where you work.

Zeeco, Inc. 22151 E. 91st St., Broken Arrow, OK 74014 USA

+1-918-258-8551 [email protected] zeeco.com©Zeeco, Inc. 2014

burners • flares • thermal oxidisers • aftermarket parts, service & solutions • rental equipment

meet the Zeeco team at the AFPM Reliability & Maintenance Conference, booth #731

Standard ZEECO® Vertical Thermal Oxidiser for Gas Processing Applications

Zeeco_PTQ Gas_Mar.indd 1 3/10/14 4:51 PMzeeco.indd 1 11/03/2014 09:10

http://www.digitalrefining.com/166,sponsors,Zeeco_Inc.html

Eliminating the Claus furnace

Fuel sources are becoming increasingly sour and society is increasing pressure on environ-

mental regulatory agencies to improve emissions standards in a desire for a cleaner environment. As a result, producers and proces-sors in the gas processing, refining, petrochemical and other industries have a growing focus on sulphur removal technologies that not only will improve the quality of their product streams but will also enable them to comply with the more stringent environmental regu-lations being imposed by governmental agencies. Since the price of sulphur is not sufficiently high for sulphur recovery processes to be profitable, the value of a sulphur recovery unit (SRU) is defined by the lost opportunity cost associated with downtime of the associated hydrocarbon processing unit as well as by its cost of opera-tion. Then, for the producers and processors, it is important to under-stand the various sulphur removal and recovery technologies available so that the most reliable, lowest cost processes can be selected and used in their facilities.

Over the past decade, researchers at Phillips 66 developed a catalytic combustor that can be used to replace the burner and thermal reactor in Modified Claus units and improve the operations of the SRU.1-9 This new approach to the Claus sulphur recovery process was presented at the Laurance Reid Gas Conditioning Conference in 2009.10 Since that time, GTC Technology has acquired the exclu-sive licensing rights to the catalytic combustion technology that

A novel approach to the conversion of hydrogen sulphide to elemental sulphur is expected to extend the economic range for Claus plants

MATT THUNDYIL, SAMEER PALLAVKAR, RAMIRO VAZQUEZ and DAVID SEEGER GTC Technology US, LLC

replaces the conventional Claus burner and thermal reactor and markets it under the name GT-CataFlame. GTC markets the complete process of GT-CataFlame integrated with the downstream Claus converters under the name GT-Spoc (Sulphur Partial Oxidation Catalysis). The ultimate design of GT-Spoc is a single vertical vessel that contains all the components of GT-CataFlame, followed by the Claus converters and sulphur condensers.

Background: conventional ClausThe Claus process was patented by Carl Friedrich Claus in 1883, and introduced in 1936. The Modified Claus process is the most successful commercial method for sulphur recovery. In this framework, the reaction of H2S with oxygen is separated into two stages: (1) a highly exothermic thermal stage where approximately 50% of the H2S is converted to elemental sulphur and, of the remaining H2S, a third of it is converted to SO2; and (2) a moderately exothermic cata-lytic stage where the remaining H2S

in the gas stream reacts with the rest of the SO2 to produce more elemental sulphur. The reactions are reversible, and conversions are highly dependent on temperature, sulphur content and moisture content. To achieve sulphur conver-sions greater than 60-70%, the thermal stage is followed by sulphur condensation and separa-tion which is followed by reheating upstream of a catalytic stage oper-ated at temperatures higher than the sulphur dew point. Additional catalytic stages may be added to increase the efficiency of sulphur removal.

The flame stability of the combus-tion section is a critical parameter in Claus operations. At a H2S content of above 55%, the acid gas can be sent directly to the furnace. Between 30–55% H2S, the acid gas or combustion air (or both) may need to be preheated. At concentra-tions below 30% H2S, the Claus unit operates in a ‘split flow’ mode with preheat and, as the H2S content drops below 10%, fuel gas may need to be added.

A two-stage Claus unit can deliver 90–95% sulphur recovery efficiency, with a three-stage config-uration delivering 95–98% recovery. The tail gas is generally sent to an incinerator if 96–97% sulphur recovery efficiency is acceptable. If sulphur recovery in the 99–99.5% range is required, tail gas opera-tions based on a continuation of the Claus reaction under sub-dew point is generally undertaken either on a solid bed, or in the liquid phase. If sulphur recovery efficiencies of 99.9% are required, the sulphur in the tail gas is generally converted


If sulphur recovery efficiencies of 99.9% are required, the sulphur in the tail gas is generally converted to H

2S by

hydrogenation and hydrolysis

gas gtc (original).indd 1 06/03/2014 12:33


near-equilibrium H2S conversion and sulphur selectivity in one tenth of the volume used by a conven-tional Claus burner and reaction furnace.10,11 The catalyst also contains components that eliminate classic Claus catalyst deactivation mechanisms of sulphur poisoning and coke deposition of the catalyst in the first Claus converter during normal sulphur recovery operation and start-up/shutdown using fuel gas.

A simplified process flow diagram for GT-CataFlame integrated with the waste heat boiler (WHB), first Claus converter and sulphur condensers is shown in Figure 1. This diagram shows an expected application where the catalytic combustor is installed in place of a poorly performing Claus furnace.

In GT-CataFlame the air and the acid gas are preheated to approxi-mately 220°C (428°F) before they are mixed. The gases are blended in a specially designed chamber to thoroughly mix the two gases upstream of the catalyst reactor bed. The gas contacts the front face of the reactor at a specific actual velocity, passes through the catalyst bed in under one second, and immediately contacts the WHB. In the reactor, approximately 70% of the H2S is converted to sulphur and H2 or H2O, 10% is converted to H2O and SO2, and most of the hydrocar-bons are converted to H2, CO and H2O, and any ammonia is converted to N2 and H2. The gas exiting the WHB may be sent to a sulphur condenser where elemental sulphur is removed; after the condenser, the gas may be re-heated to pass into the first Claus converter. Alternately, the gas exiting the WHB may pass directly into the first Claus converter and the previously mentioned sulphur condenser and re-heater may be eliminated; pass-ing directly from the WHB into the first sulphur condenser the process effectively ‘jumps’ the Gamson and Elkins curve and 90% sulphur recovery occurs after the first Claus converter.12 The process that includes sending the gas directly from the WHB into the first Claus converter is part of the patents that

to H2S by hydrogenation and hydrolysis. Then the H2S is captured and recycled to the Claus unit.

Modified Claus units are chal-lenging to operate reliably, and are particularly prone to problems during start-up and shutdown. For example, the furnace is usually started up and shut down using a fuel gas stream as the fuel, rather than the acid gas. When the feed shifts from acid gas to predomi-nantly fuel gas, with an associated shift in oxidant ratio (air ratio), higher temperatures and soot formation often may result, among

other undesirable consequences. The soot has a tendency to foul the Claus catalyst downstream of the furnace.

The operating cost associated with sulphur removal for a Claus unit with tail gas clean-up is in the $100/t range when considering utilities and maintenance costs. The sulphur that is recovered is gener-ally bright yellow and preferred in the marketplace.

GT-Spoc: process descriptionGT-Spoc uses a patented, durable catalyst in a ‘short-contact-time’ reactor, GT-CataFlame, to achieve

Reheat

Catalyst bed

Condenser

Sulphur pit

Condenser

GT-CataFlame

WHB

Acid gas

Air

Boiler feed water

HP steam

LP steam

Figure 1 Simplified process flow diagram of GT-CataFlame integrated with the first Claus converter and sulphur condensers

Acid gas

Air

GT-CataFlame reactor

Equivalent conventional Claus furnace

Waste heat recovery

Waste heat recovery

Figure 2 A conventional Claus furnace and a GT-CataFlame designed for the same application showing the relative size difference



chimney-type separation tray where the liquid sulphur exits the side of the unit and the gas passes down through the chimney. After this molten sulphur separation tray, the remainder of the process is the

same as a conventional SRU except that it is arranged in a vertical confi guration.

The process that includes remov-ing the fi rst sulphur condenser (the one located after the WHB and

Stream Acid gas from amine unit Sour water stripper gas Combustion airMass fl ow, kg/h 4050 730 10 900Flow rate, Nm3/h 2671 675 8500 Composition, kg-moles/h C

1 0.48 0 0

C2 0.72 0 0

C3 1.19 0 0

C4 0.60 0 0

CO2 1.19 0 0

H2O 2.86 6.00 5.4

H2S 112.21 12.00 0

Hexene 0 0.04 0Phenol 0 0.02 0N

2 0 0 295.6

O2 0 0 78.5

NH3 0 12.00 0

Conditions for comparison of GT-CataFlame and a Claus furnace

Table 1

HP steam

Boiler feed water

Boiler feed water

Boiler feed water

LP steam

Liquid sulphur

LP steam

Condensate

LP steam

Liquid sulphur

Tail gas

Preheated acid gas

Preheated air

GT-CataFlame catalyst

Claus catalyst

Claus catalyst

Figure 3 Ultimate vertical design of GT-Spoc with two Claus converters in series

are included in GT-Spoc. The gas path after the fi rst Claus converter is the same as that of a conven-tional Claus SRU.

A conventional Claus furnace and a GT-CataFlame reactor designed for the same acid gas application are shown for comparison in Figure 2; the conditions used to design both units are summarised in Table 1. In the fi gure, the furnace and GT-CataFlame are drawn to the same scale which shows the rela-tively larger size requirement for the furnace. This size difference saves signifi cantly on footprint and the amount of refractory material required. The GT-CataFlame refrac-tory volume is 1/40th that of the refractory volume required for the furnace. This large reduction in refractory is due in part because the refractory lining the mixing section is only for protection and good engineering practice.

The ultimate design intended for GT-Spoc is a vertical arrangement; that is, GT-CataFlame is integrated with the WHB and the downstream Claus unit equipment, all combined in a single vertical tower. A simpli-fi ed diagram of the vertical GT-Spoc with two Claus converter stages is shown in Figure 3. The acid gas and air are fi rst pre-heated as they were in the horizontal arrangement described above. The preheated gases are then combined at the top of the unit in the mixing chamber. The well-mixed gas then travels downwards to the catalyst where the H2S is converted to sulphur. Next, the gas passes to the fi rst Claus converter without fi rst passing through a sulphur condenser. In the vertical GT-Spoc arrangement, the sulphur condenser located between the WHB and the fi rst Claus converter is removed. The WHB operating temperature is adjusted so that the gas out of the fi rst Claus converter remains above the sulphur dew point temperature. The gas exits the WHB and passes directly to the fi rst Claus converter bed. After the Claus converter, the gas is cooled in the fi rst sulphur condenser and molten sulphur forms as it does in the conventional Claus SRU. The molten sulphur collects on a



• The least amount of unwanted byproduct formation, primarily COS and CS2• The destruction of ammonia • Formation of H2 and CO, rather than any unwanted hydrocarbon byproducts.

The results shown in Figure 4 indi-cate the performance of a particular catalyst formulation that was tested for an extended period in the unit. During this testing period, air-to-H2S ratios and gas velocities were being evaluated so, throughout the duration of the testing, these param-eters were varied to observe the effects and to arrive at optimum operating conditions. However, even though conditions were changed, H2S conversion and selec-tivity towards forming elemental sulphur remained high throughout the period. At certain points, the data reflect operating conditions that achieve better conversion and sulphur yield so those operating conditions will be used for commer-cial unit design and operation. Throughout the test period, the range of gas composition was approximately 80-82 mol% H2S, 11-12 mol% CO2, 5-6 mol% water, and 1 mol% hydrocarbon content; the remaining 1-2 mol% is nitrogen and the typical analytical measure-ment error.

The results indicate that over 80% of the H2S was converted with approximately 69-70% yield of elemental sulphur; the remainder was unreacted H2S and SO2, and less than 1% formation of COS and CS2 combined. In addition there was an average 7% yield of H2. This performance was better than that of a conventional Claus furnace where 50–70% of the inlet H2S is converted and the elemental sulphur yield is 40–65% of the inlet sulphur.

AdvantagesBased upon the results of the test-ing that have been presented here and previously,10 several process and economic advantages have been identified from having a short contact time catalytic combustor versus a flame reaction furnace.

Process advantages• Reactants are pre-mixed prior to

showed significant improvement in hydrogen and sulphur yields over what was typically observed in refinery Claus furnaces. Higher sulphur and hydrogen yields were believed to result from direct oxidation and subsequent splitting of H2S at high reaction temperatures.11 The results were gathered with no hydrocarbons in the inlet gas, therefore the hydrogen must come from NH3 and H2S.

Operating resultsThe researchers from Phillips 66 undertook a project to demonstrate the catalytic combustion section of GT-Spoc, GT-CataFlame. A pilot unit installed at a refinery location in the US operated intermittently for more than a year. Testing was performed to identify the operating parameters and catalyst formula-tion that resulted in:• The highest H2S conversion to sulphur• Producing the necessary amount of SO2 for the downstream Claus converters

upstream of the first Claus converter) is patented by Phillips 66 and licensed by GTC. This step has the advantage of reducing the number of pieces of equipment, the overall SRU footprint, and reducing the cost of the SRU; this configura-tion may be applied to a conventional Claus SRU with or without a GT-CataFlame as well as in the vertical GT-Spoc SRU.

GT-Spoc: operating dataPhillips 66 conducted a large number of tests on a catalytic combustor that could be used to replace the conventional Claus furnace with the goal of providing improved operation of Claus SRUs. After extensive development, the inventors arrived at a robust catalyst formulation and a combus-tor design that delivered results which could be used to model and design commercial facilities. The summary results presented in Table 2 are from tests using the catalytic combustor compared to measurements taken at the SRUs of four refineries. The results

60

100

90

80

70

50

40

30

20

10

H2 c

oncentr

ati

on

H2S

conve

rsio

nS

ulp

hur

yeild

, %

00 100 200 300 400 500 600 700 800 900

Elapsed time, hrs

Sulphur yieldH2S conversion

Outlet H2

Figure 4 Results of H2S conversion, sulphur yield and hydrogen concentration in the

outlet gas for 900 hours of testing at various operating conditions in a GT-CataFlame unit

Unit GT-Spoc Refinery A Refinery B Refinery C Refinery DNH

3 present? Yes Yes Yes Yes No

% S yield 74.0 63.4 48.1 54.9 68.2% H

2 yield 7.0 6.3 3.5 5.4 4.2

Laboratory results for GT-Spoc catalyst compared with refinery Claus reaction furnace sulphur and H

2 yields

Table 2


fast returns

A Century of Innovationin the Oil and Gas Industry1914 - 2014

Quickly maximize the value of natural gas assets with NGL recovery solutions from UOP.

Get cash flow and profits sooner with hydrocarbon management solutions fromUOP, including the UOP Russell line of complete, modular, factory-built systemsthat are designed to have you up and running six months quicker than otheroptions. The reduced on-site installation time, proven designs and pre-fabricatedsolutions can deliver an earlier startup at a lower cost than stick-built facilities.Or choose the UOP-Twister® Supersonic Gas Separation System, with a smallfootprint and low maintenance that’s often ideal for remote and offshore facilities.Or get maximum NGL/LPG recovery — an industry-leading 99+% — with solutionsfrom UOP and Ortloff Engineers, Ltd. In short, whether you need fast, compact,efficient or all of the above, UOP has a solution for your NGL-recovery needs.

For more information about UOP Russell solutions, visit www.uop.com/uoprussell orvisit www.uop.com to learn about all of the UOP hydrocarbon management solutions.

© 2014 Honeywell International, Inc. All rights reserved.

UOP-Russell-Fast-Returns-Ad-Full-Page_UOP Ad 8.5x11 UOP6467 1/10/14 3:22 PM Page 1

uop.indd 1 04/03/2014 10:48

http://www.digitalrefining.com/149,sponsors,UOP___A_Honeywell_Company.html


re-heater that is located down-stream of the WHB may also be eliminated, so that the gas stream exiting the WHB flows directly to the first Claus reactor stage.

COS and CS2 reduction and

implications for Claus converter bed designThe best way to minimise the effects of hydrocarbon contamina-tion of acid gases in any SRU is to prevent hydrocarbon absorption in the acid gas removal system (slug and drop catching drums, aerosol droplet removal, and lean amine/sour gas differential temperature control) and mitigate hydrocarbon accumulation (rich flash drums with skimming and gas removal, carbon filtration, and regenerator reflux purge).

Despite these efforts, hydrocarbon may get into the acid gas, leading to the undesired byproducts COS and CS2. Nominally, the byproducts result from reacting after the flame mixing zone where hot, oxygen-free flame byproducts can mix with

ratios, making switching nearly seamless • COS and CS2 formation are significantly reduced• The sulphur product contains only 25-33% of the dissolved H2S/H2Sx of normal Claus sulphur.

Economic advantages• Significant reduction in refractory lining due to smaller size of GT-CataFlame • Small reactor volume and reduc-tion in byproducts allows for design changes reducing overall unit footprint• Design changes eliminate inter-connecting piping and sulphur rundown piping and equipment• Because GT-CataFlame is smaller than a typical furnace, there is less refractory and catalyst mass to heat up or cool down; this in turn reduces the time required for shut-down or start-up• The Claus plant can be constructed in a vertical orientation where the unit is self-draining• The first sulphur condenser and

passing through the GT-CataFlame reactor. The reaction is uniform through the cross section of the reactor eliminating the problems of post-combustion mixing for contaminant destruction• Close-coupling of the catalyst zone and waste heat boiler (WHB) improves overall sulphur yield and reduces air requirement due to rapid reaction quenching• No flames, fire-eyes, nor burner management systems: a small retractable burner or access to a hydrogen containing fuel is all that is needed to trigger the catalyst at start-up• Fuel gas oxidation for warm-up takes place at 25% air stoichiometry, eliminating oxygen breakthrough to downstream Claus converter beds during start-up and shutdown• Low molecular weight hydrocar-bons are converted to reduction tail gas unit-friendly H2 and CO by catalytic partial oxidation. Soot formation is virtually eliminated. Air/fuel ratios for fuel gas oxida-tion closely resemble acid gas/air

www.eptq.com PTQ Q1 2014 131

3 Respini, Jones, Spanu, Sesselego, Avoiding foul play, Hydrocarbon Engineering, Nov 2006.

Matteo Virzi is Senior Technology Manager with ISAB Priolo refinery in Italy. He is an expert in distillation and thermal conversion processes. With more than 20 years of experience in technology, operations and automation, he holds a degree in chemical engineering from the University of Palermo. Email: [email protected]

Marco Respini is a Senior Technology Expert with Baker Hughes Downstream Chemicals, specialising in refinery and petrochemical process improvements in fouling control. He has 15 years of refining experience and is currently involved in developing new technologies for improving refinery conversion processes. With extensive experience in asphaltene related problems in oil production and refining, he is an inventor of five US patents and has published 10 technical papers and seven conference papers on visbreakers and heavy fuel oil stability problems. A graduate of Milan University with a degree in industrial chemistry, he has been a Research Fellow in the field of organometallic catalysts and is a registered professional chemist in Italy. He is also a member of ACS and NACE. Email: [email protected]

which provided refinery personnel with the right information to be able to constantly keep the unit at the best process severity for any processed feed.

VisTec is a mark of Baker Hughes Incorporated.

References1 Petralito G, Respini M, Achieving optimal visbreaking severity, PTQ, Q1, 2010.2 A Phase Separation Kinetic Model for Coke Formation, Preprints ACS, Div. Pet Chem, 38, 428-433, 1993.

result in better handling of fouling, and at the same time increase HVGO directly (less gasoil recycled as wash oil). Figure 17 shows the trend of wash oil rate.

The decrease in wash oil rate was continuously optimised. The HVGO was analysed on a daily basis with respect to the level of contaminants, using the VCI tech-nique that measures the coke particles entrained within the HVGO (thus not removed by wash oil).

Also, the rate of antifoulant on the wash oil was optimised and increased when needed.

ConclusionsThis article describes an example of high conversion visbreaking with increased run length when compared to typical visbreaker run lengths.

This successful result was made possible by coupling Baker Hughes VisTec anticoke/antifoulant treat-ments with monitoring technology,

A controlled decrease in wash oil can result in better handling of fouling, and at the same time increase HVGO directly

Best Valves

since 1867

baker hughes.indd 8 13/12/2013 11:38gas gtc (original).indd 5 11/03/2014 19:05

http://www.digitalrefining.com/169,sponsors,OHL_Gutermuth_Industrial_Valves.html


Start-up and shutdown operationalimprovementsMost of the time, starting up a Claus plant is spent heating up the large amount of refractory in the reaction furnace chamber and catalyst in the catalyst beds. The GT-CataFlame reaction section is very compact and in close proximity to the waste heat boiler entrance, so there is a mini-mum amount of refractory to be dried and heated to operating temperature.

The small amount of catalyst and support materials in the reaction section is easily heated up by a small preheater such as a Stackmatch, eliminating one of the biggest problems in starting up or recovering from a shutdown —relighting the main burner. Laboratory testing has shown that hydrogen rich gases (~40% H2) and air mixtures can be catalytically ignited by the catalyst, potentially bypassing the need for any type of fired preheater.

Minimising the formation of contaminants such as COS and CS2 and operating the first catalyst bed directly after the WHB makes it possible to reduce the amount of Claus catalyst needed, speeding up both the shutdown process (sulphur removal or heat soak) and the start-up process (warm-up). Elimination of the sulphur condenser following the WHB and subsequent reheater eliminates cool-down or warm-up of equip-ment and interconnecting piping, again speeding up the shutdown or start-up process.

When heating up or cooling down a Claus plant, fuel gas firing at near stoichiometric ratios is used, especially when Claus cata-lyst is sulphur laden. Air-to-fuel gas ratios for most natural gas streams range from 9.5-10:1, while air to acid gas ratios are typically in the 1-2.5:1 range, depending on acid gas concentra-tion. This makes for difficult switching from one feed gas to another. At stoichiometric ratios, the adiabatic flame temperature (~3500°F, 1925°C) for most fuel gases is well above the temperature limits of reaction furnace refractory (~2800°F, 1535°C) and ferrules,

The GT-CataFlame demonstration unit confirmed the results obtained in the laboratory.

Turndown Laboratory testing showed that operations were stable down to 25% turndown when operating only with acid gases. However, the laboratory apparatus was capable of running with any mixture between pure natural gas and pure H2S.

Demonstration plant data were run at as low as 12% turndown with minimal effect on perfor-mance. Because of the ability to operate with nearly any ratio of fuel gas to acid gas, even lower turndown rates can be achieved.

Air demand and related operationalimprovementsWhile the overall recovery of sulphur will fluctuate significantly with air-to-acid gas ratios in both

the Claus and GT-Spoc processes, the GT-CataFlame catalyst’s selec-tivity for sulphur does not vary significantly over a wide range of air-to-H2S ratios.

The combination of pre-mixing before reaction and close coupling of the reactor to the WHB aids both hydrogen formation reactions (partial oxidation and dissociation) and inhibits the main hydrogen consumption reaction (recombina-tion of H2 and S). As flow rate increases in the unit, improved heat transfer due to higher Reynolds number and thermal conductivity of the gas (due to higher hydrogen content) appear to aid in inhibiting recombination. This translates to more capacity with less overall pressure drop.

unburned hydrocarbons, for instance CH4 and S2, in the acid gases as in a typical 1/3-2/3 split-flow furnace for leaner H2S content gases.14 Since the gases flowing to the GT-CataFlame reactor are pre-mixed and pre-heated, the temperature is hotter than that of a furnace (approximately 2200-2300°F, 1200-1260°C), and since the time from reactor to WHB is short, plant data show a significant reduction in the amount of byproduct COS and CS2 when compared to a typical Claus unit.

A study of 24 gas plants with a broad range of inlet H2S concentra-tions in the gas passing to the Claus furnace showed that the COS and CS2 concentration in the gas exiting the WHB was in the range 0.06-1.7 mol% for COS and 0.01–1.1 mol% for CS2.

13 At the optimised operat-ing conditions for the GT-Spoc pilot unit, 0.12–0.18 mol% COS and 0.36–0.48 mol% CS2 were measured in the gas leaving the catalytic combustor.

Also, since the formation of COS and CS2 is significantly decreased, and the first Claus converter condi-tions will be hotter than a conventional Claus SRU design, the Claus converter following the reac-tor can be much smaller, helping to reduce the Claus catalyst bed size and allowing for a more compact footprint. In GT-Spoc, the first Claus converter may be hotter than in a conventional Claus design if the gas out of the WHB flows directly to the Claus converter without being cooled in a sulphur condenser.

Acid gas concentration requirementStable operation of the GT-CataFlame catalyst was demon-strated at acid gas concentrations as low as 25% H2S using only pre-heated air and pre-heated acid gas. In addition, a wide range of laboratory tests were conducted on lean H2S streams containing CO2 or nitrogen diluents and typical light hydrocarbon components. Sulphur yields varied with H2S concentra-tion, but the yields were still close to equilibrium computations, unlike lean acid gas Claus units that produce little to no sulphur from the thermal stage.

In general, the more frequent the shutdown and start-up of a Claus unit, the more problems can be expected



Capital cost comparisonAn independent engineering firm performed a cost study that compared the capital cost of a two-stage GT-Spoc unit to a conventional two-stage Claus unit with a furnace. The GT-Spoc SRU was designed in the single, vertical tower arrangement shown in Figure 3. The amine unit, the tail gas unit (TGU) and the sulphur pit were assumed to be similar for both units and were not included in the capital cost estimate. Therefore any cost advantage of the effects of reduced duty on the TGU and less H2S and H2Sx dissolved in the sulphur are not included in the comparison. The cost study was prepared using the standard engi-neering cost factors based on US Gulf Coast prices.

Design basisThe composition of the acid gas flowing to the GT-CataFlame or to the Claus furnace is shown in Table 4. For purposes of presentation in the table, the acid gas and sour water stripper (SWS) off-gas are added together and reported as the composition of one stream, as they would be in the case of the feed gas to the GT-Spoc SRU. However, in the case of a conventional Claus stream, the SWS off-gas is added to the burner and the acid gas is oper-ated in split flow.

The flow rate of the combined acid gas and SWS off-gas is 42.9 Nm3/h (3.8 MMscfd) and the sulphur production rate is approxi-mately 100 t/d.

Cost comparison resultsThe results of this cost estimate indicated that the capital cost of a GT-Spoc unit is approximately 20% less than that of a conventional Claus unit. A few major differences between the two processes are the reasons for the savings and they are:• The GT-CataFlame reactor is approximately 20% of the volume of an equivalent Claus furnace• GT-CataFlame requires much less refractory (as little as 1/40th the amount required for a conventional Claus furnace)• The sulphur condenser located

Reduction of H2S and sulphanes in

produced sulphur Experimental data from laboratory tests demonstrate that the sulphur produced in the GT-Spoc vertical arrangement contains less dissolved H2S and H2Sx because all the sulphur from both stages will only be exposed to the H2S partial pres-sure that would normally be found at the second condenser of a conventional Claus unit, which is significantly lower than at the first condenser following the conven-tional thermal stage. The results in

Table 3 show the difference in H2S and H2Sx content as the sulphur from the sulphur condenser imme-diately following GT-CataFlame contains five to six times more dissolved H2S and H2Sx as sulphur from the condenser that follows both GT-CataFlame and the first stage Claus converter.

requiring the incorporation of steam or nitrogen diluent to hold flame temperatures to a safe oper-ating envelope.

Also, near-stoichiometric flames have a tendency to form soot which can rapidly plug catalyst beds downstream of the flame. GT-CataFlame catalyst can convert fuel gas via partial oxidation (down to as low as 25% of full stoichio-metric) air, close to the ratios used for acid gas operations. The catalyst has components that resist the formation of coke on the catalyst surface which translates into less production of soot.

Firing slightly over stoichiometric can cause unwanted catalyst exotherms, damaging both catalyst and vessels. Since the catalyst does not have to operate near stoichiometric air-to-fuel ratios, damaging exotherms on Claus cata-lysts are eliminated.

In general, the more frequent the shutdown and start-up of a Claus unit, the more problems can be expected. During the small demon-stration plant’s long term catalyst testing, the unit was shut down and started up over 30 times. Through all of the start-ups and shutdowns, the GT-CataFlame catalyst retained its original activity.

Conditions:Air/H

2S 2.4

H2S flow 750 cm3/min After GT-CataFlame After GT-CataFlame and one Claus stage

% Sulphur yield 72.3 85.8H

2S in sulphur, ppmw (FTIR) 308.4 60.3

H2S

x in sulphur, ppmw (FTIR) 340.1 54.2

Reduction of H2S and H

2S

x in sulphur passing through Claus catalyst

Table 3

Component Mole % Component Mole %CO

2 0.20 nC

4 0.00

H2S 67.80 nC

5 0.02

C1 0.11 nC

6 0.00

C2 0.09 H

2O 19.80

C3 0.02 NH

3 11.94

iC4 0.00 H

2 0.05

Acid gas composition used for a capital cost comparison of GT-Spoc and a conventional Claus SRU

Table 4

The capital cost of a GT-Spoc unit is around 20% less than that of a conventional Claus unit



Enhanced Synthesis Gas Production.6 US 7,138,101 Two Stage Catalytic Process for Recovering Sulfur from an H

2S Containing

Gas Stream.7 US 7,226,572 Compact Sulfur Recovery Plant and Process.8 US 7,326,397 Catalytic Partial Oxidation Process for Recovering Sulfur from an H

2S

Containing Stream.9 US 7,357,908 Apparatus and Catalytic Partial Oxidation Process for Recovering Sulfur from an H

2S Containing Gas Stream.

10Keller A, Is there “direct oxidation” of H2S

to sulfur?, Laurance Reid Gas Conditioning Conference, Feb 2009.11 Partial Oxidation of H

2S to sulfur, a Claus

Alternative, British Sulphur Conference, Oct 2003.12 Paskall H G, Sames J A, Sulphur Recovery, 13th Edition, 2010, 1-14.13 Paskall H G, Sames J A, Sulphur Recovery, 13th Edition, 2010, 2-80.14Borsboom H, Clark P, Goar B G, New insights into the Claus thermal Stage, chemistry and temperatures, Laurance Reid Gas Conditioning Conference, Feb 2009.

MattThundyil is Manager of Sulfur Business for GTC Technology US, LLC, in Houston, Texas. He has over 15 years’ experience in the petrochemical and energy industry and is the inventor on a number of patented separations technologies. He holds a degree in chemical engineering from the Indian Institute of Technology (IIT), Madras, India, and a PhD from The University of Texas, Austin, in advanced separations technology for natural gas processing. SameerPallavkar is a Sulfur Process Engineer with the Sulfur Business Group at GTC Technology US, LLC’s Houston office. During his PhD program at Lamar, he worked as a research assistant in the Department of Chemical Engineering, and was assigned to the Microwave Reaction Engineering Laboratory as a research candidate. He holds a degree in chemical engineering from the University of Mumbai, Bombay, and a Master’s from Lamar.Ramiro Vazquez is a Licensing Manager, Sulfur Business Unit, at GTC Technology US, LLC’s Houston office, where he is responsible for technology licensing and sales. He holds a degree in chemical engineering from the University of Nuevo Leon, Monterrey, Mexico, and a Master’s in chemical engineering from the University of Houston. David Seeger, Senior Scientist for GTC Technology US, LLC’s Sulfur Group in Houston, holds a BA degree in chemistry from St. Thomas University (St. Paul, MN) and MA and PhD degrees in chemistry from the University of Michigan (Ann Arbor, MI). He is the inventor of a selenium removal process for FGD and has led research efforts to develop and improve processes for HAP, FGD and sulphur removal. He is one of the founders of CrystaTech.

after the WHB and before the first Claus converter in a conventional Claus SRU is eliminated in the vertical GT-Spoc SRU design• The first reheater that follows the first sulphur condenser in a conven-tional Claus SRU is also eliminated as it is unnecessary in GT-Spoc• Interconnecting piping is reduced in GT-Spoc as the stages are close coupled in a vertical arrangement as compared the conventional Claus SRU.

SummaryConventional Modified Claus tech-nology with a free-flame thermal section is well known to have a number of operating challenges associated with the reactions that occur in the flame. An alternative, proven approach is presented here, where the free-flame thermal section is replaced with a catalytic combustor that utilises a durable catalyst that enables sulphur partial oxidation catalysis. This change at the front end of a conventional Claus process results in significant process and economic advantages, including improved sulphur yield, reductions in COS and CS2 forma-tion, reduced impact of hydrocarbons on air demand, improved start-up and shutdown operations, and reductions in the Claus converter catalyst bed sections. Overall, GT-CataFlame allows upgrades of aging Claus burners and enables a 20–30% reduction in capital costs associated with Claus plants, while reducing the tail gas unit load due to its higher conversion efficiencies.

GT-SPOC and GT-CataFlame are marks of GTC Technology US, LLC.

References1 US 6,403,051 Recovery of Sulfur from H

2S

and Concurrent Production of H2 Using Short

Contact Time Reactors.2 US 6,579,510 SPOX Enhanced Process for Production of Synthesis Gas.3 US 6,800,269 Short Contact Time Catalytic Sulfur Recovery System for Removing H

2S from

a Waste Gas Stream.4 US 6,946,111 Short Contact Time Partial Oxidation Process for Recovering Sulfur from an H

2S Containing Gas stream.

5 US 7,122,170 Catalysts for GT-SPOC™


of Heavy Hydrocarbonaceous Feedstock, European Patent Application, EP 1754770A1, Research Institute of Petroleum Industry. 8 Bahmani M, Sadighi S, Mohaddecy S R, Mashayekhi M, Hydrocracker parametric sensitivity study, PTQ, Q2 2009.9 Sadighi S, Mohaddecy S R, Ghabouli O, Rashidzadeh M, Optimisation of product yield and coke formation in a RFCC unit, PTQ, Q2 2010.Sepehr Sadighi is Project Manager, Catalysis and Nanotechnology Division, Catalytic Reaction Engineering Department, Research Institute of Petroleum Industry (RIPI), Tehran, Iran. He holds a PhD in chemical engineering from Universiti Teknologi Malaysia. Email: Sadighis @ripi.irReza Seif Mohaddecy is Project Manager, Catalysis and Nanotechnology Division, Catalytic Reaction Engineering Department, RIPI. He holds a MS in chemical engineering from Sharif University of Technology. Email: Seifsr @ripi.irKamal Masoudian is Project Manager, Catalysis and Nanotechnology Division, Catalyst Characterization and Evaluation Department, RIPI. He holds a BSc in chemical engineering. Email: masoudiansk @ripi.ir

References1 Bhaskar, M, Valavarasu G, Balarman K S, Advantages of mild hydrocracking FCC feed - a pilot plant study, Petroleum Science and Technology, 21, 9 & 10, 1439-1451, 2003.2 European patent application, EP 1754770A1, Process for Hydroconverting of a Hydrocarbonaceous Feedstock, Research Institute of Petroleum and NTI company, 2007.3 Mohaddecy S R, Zahedi S, Sadighi S, Bonyad H, Reactor modeling and simulation of catalytic reforming, Petroleum & Coal, 48 (3), 28-35, 2006.4 Lee R, Leunenberger E, Powell R, Optimizing the cat feed hydrotreater/FCCU complex with detailed simulation tools, Process Technology Update, Desulphurization Process, Word Refining, Jul/Aug 2001.5 Dean R R, Mauleon J L, Combined Fluid Catalytic Cracking and Hydrocracking Process, US patent No: 4426276, 17 Jan 1984.6 Talman J, Jonsgma B, Thamprajmachit B, Cackett S, Wijk R, Synergistic integration of FCC and hydroprocessing facilities for upgrading bottom of the barrel, Asia Pacific Refining Technology Conference, 2001, Kuala Lumper. 7 Kadiev K M, Mezhidov V K, Zarkesh J, Masoudian S K, Process for Hydroconverting

60

50

40

30

20

10

Die

sel in

cre

ase

, %

0Simple series Series Parallel Residue

upgradingIntegration cases

39.89% 41.27%37.53%

53.69%

Figure 7 Comparison to base case of growth in diesel production

HCR feed Ni+V, ppmwt Sulphur, wt% IBP, °C FBP, °CBase 2.98 1.222 309 516.4Simple series 2.98 1.222 309 516.4Series 2.98 1.222 309 516.4Parallel 3.18 1.323 285.5 516.3Residue upgrading 3.18 1.323 285.5 516.3

Feed quality of HCR unit in base case and integration strategies

Table 6

RFCC feed Ni+V, ppmwt Sulphur, wt% IBP, °C FBP, °CBase 14.723 1 264.6 592.6Simple series 14.723 1 264.6 592.6Series 14.615 1.006 264.6 599.5Parallel 14.874 0.9667 267.3 592.6Residue upgrading 13.734 1.088 267.3 592.1

Feed quality of RFCC unit in base case and integration strategies

Table 7

www.eptq.com Revamps 2013 63

processes might increase in the future, and the fouling of units might become a more serious concern, potentially impacting a large number of production days. By regular application of online cleaning, the unit can always be operated under clean conditions. In addition, the vast majority of mechanical work, which takes up a significant number of days during unit shutdowns, can be replaced with an online cleaning process that requires no opening of equipment or man entry, and can be carried out in as little as 24 hours oil-to-oil.

AcknowledgementSpecial thanks to all the Lotos team for the support and valuable assistance provided during the entire time that ITW Online Cleaning has been applied on site.

Mariusz Hołowacz has been Crude Oil Distillation Complex Manager with Grupa Lotos in Gdansk, Poland, since 2008 and, since 2010, Manager of the new distillation unit 120.Rafał Zaprawa was VDU and propane deasphalting unit Shift Team Manager with Grupa Lotos in Gdansk, Poland, from 2001-2008, then Deputy Manager of the crude oil distillation complex. Marcello Ferrara is the Chairman of ITW. With 27 years’ experience in the petroleum business, including oil exploration and production, refining, petrochemicals, transportation, and energy production, he holds a PhD in industrial chemistry and international patents for new processes and additive compositions for environmental control and for improving petroleum/petrochemical processes. Email: [email protected]

market considerations and by the downtime required for a mechani-cal cleaning turnaround. With a downtime of 15-20 days, it is much more economical to run the units under non-optimised conditions rather than lose production. This, however, results in energy losses, giveaway and capacity reduction, which negatively impact unit economics.

The introduction of online clean-ing, by cleaning the unit in 24 hours on an oil-to-oil basis, allows for the recovery of losses and the operation of units under improved and more reliable conditions. Furthermore, turnarounds can be avoided or rescheduled with reduced downtime.

In the case of turnaround improvement, an additional reduction in downtime can be achieved by applying ITW’s improved degassing/decontamina-tion technology.

ConclusionThe results of ITW Online Cleaning have opened up new possibilities for Lotos, whereby online cleaning can be applied during a plant run, to recover a unit’s performance and improve the level of operational excellence, and in preparation for a turnaround, to reduce downtime and the turnaround’s scope of work.

These options are increasingly important, as the amount of oppor-tunity crudes the company

Figure 7 HGO — crude exchanger

Heavy crude Oil

Atmospheric Distillation

Vacuum Distillation

Coker & Visbreaker Feed

Fluidized Catalytic Cracker

Bitumen

Non-intrusive fl ow measurement

up to 400°C

Field-Proven at Refi neries

Trouble free operation at extreme pipe temperatures

No clogging, no abrasion, no pressure losses

Installation and maintenance without process interruption

Independent of fl uid or pressure

Hazardous area approved

Non-intrusive fl ow

refi nery.fl exim.com

itw.indd 5 10/09/2013 17:06ripi.indd 6 11/12/2013 20:18gas gtc (original).indd 8 11/03/2014 19:06

http://www.eptq.com/jump.aspx?a=Flexim&p=shale14

Reliable SwiSS Quality

API 618Rod load up to 1'500 kN/335'000 IbsPower up to 31'000 kW/42'100 HP

Full Range:

You get moRe than Just a PRocess gas comPRessoR

Lubricated up to 1'000 bar, non-lubricated up to 300 bar

For highest availability: We recom-mend our own designed, in-house engineered compressor valves and key compressor components

Designed for easy maintenance

We are the competent partner with the full range of services – worldwide

→ www.recip.com/api618

Reliable SwiSS Quality

API 618Rod load up to 1'500 kN/335'000 IbsPower up to 31'000 kW/42'100 HP

Full Range:

You get moRe than Just aPRocess gas comPRessoR

Lubricated up to 1'000 bar, non-

For highest availability: We recom-mend our own designed, in-house engineered compressor valves and key compressor components

Designed for easy maintenance

We are the competent partner with the full range of services – worldwide

YouR BeneFit: lowest liFe cYcle costs

gas burckhardt.indd 1 04/03/2014 10:45

http://www.eptq.com/jump.aspx?a=Burckhardt&p=shale14

Predicting corrosion rates in amine and sour water systems

Corrosion is a ubiquitous prob-lem in gas treating in the petroleum and natural gas

industries, in syngas plants, in processing unconventional gases such as shale and coal seam gas, and in numerous other treating applications. The primary impuri-ties removed in the treating process are the acid gases carbon dioxide and hydrogen sulphide. The corro-sion of equipment and piping is an inevitable consequence of removing these very gases with amines, and of handling sour water. There are other corrosive impurities that either enter in small amounts with the gas, such as HCN and oxygen, or that are produced in the amine system itself, mostly heat stable salts (HSS) derived from HCN. Corrosion rates are affected by the nature of the corrosive agent, temperature, fluid velocity, the presence of solids, and the metal-lurgy involved.1 To prevent equipment failures, mitigate risk and select optimal materials, one must be able to predict corrosion rates pertinent to the particular processing conditions. This article describes the underpinnings of a chemistry-based predictive corro-sion model built on both public and much proprietary corrosion rate data. The model includes depend-ence on ionic solution composition (speciation), fluid velocity, temper-ature, HSSs, and metallurgy.

Corrosion in alkaline systemsAlthough the concepts presented here apply equally to pH-neutral and acidic systems, these systems are not addressed because the amount of corrosion data available

A chemistry-based predictive model predicts corrosion rates in specific processing conditions

NATHAN A HATCHER, CLAYTON E JONES, G SIMON WEILAND and RALPH H WEILANDOptimized Gas Treating, Inc.

for modelling is not as extensive. The corrosive action of H2S is inher-ently different from that of CO2 in that H2S can and does form a rela-tively robust, protective iron sulphide layer on the metal surface. On the other hand, iron carbonate forms a more fragile layer, so it offers much less protection. There are several tenets embedded in the model:• The corrosive agents are acids• In and of itself, the amine (or ammonia) is not corrosive

• The iron sulphide film can protect against further corrosion• Iron carbonate also offers protec-tion but to a lesser degree• High fluid velocities physically increase corrosion rates• Higher temperature increases corrosion rates• Heat stable salts chemically exac-erbate corrosion.

Although technically incorrect, the industry continues to bandy about such terms as ‘amine corro-sion’ and ‘alkaline stress corrosion cracking’ to describe corrosion that, at the root level of chemistry, is really caused by dissolved acid

gases in various forms. For exam-ple, nearly 60 years ago, Polderman2 reported that 20 wt% MEA without acid gas was actually less corrosive to steel than pure water. As far as the corrosive agents themselves are concerned, the important parameter is the chemical activity of the dissolved acid gas species responsible for corrosion. The activity (vs concen-tration) changes with the amine type, amine concentration, acid gas loadings, the concentrations and identities of HSSs, and temperature. This may make some amine systems appear to be more suscep-tible to corrosion than others; however, the essential point is that it is the activity of the corrosive species that is of direct importance, not the type of amine per se.

The chemical species of interest are: bisulphide ion (HS–), free phys-ically dissolved H2S, bicarbonate ion (HCO3

–), and free physically dissolved CO2, all of which are oxidising agents. These species are called protonic acids because they can give up a hydrogen ion.3 Sulphide (S=) and carbonate (CO3

=) ions are also present; however, they themselves are final reaction prod-ucts and are unable to provide the hydrogen ion necessary for the oxidation of iron. Molecular hydro-gen sulphide and carbon dioxide react with iron only in the presence of water. The final distribution of molecular and ionic species is found by solving the equations of chemical reaction equilibria, atom balances, and a charge balance. The resulting set of species concentra-tions is termed the solution’s speciation.


Corrosion rates are affected by the nature of the corrosive agent, temperature, fluid velocity, the presence of solids, and the metallurgy involved

gas ogrt.indd 1 06/03/2014 13:37


test, CO2 was added to an amine solution that already contained H2S. The corrosion rate quickly rose and in less than one hour a new higher steady corrosion rate was estab-lished. The precipitated layer of iron carbonate is very fragile, so it is unlikely to adhere as tenaciously to metal surfaces as iron sulphide does. Different abilities to adhere are contained within the framework of the present model.

Managing corrosion begins primarily with protecting the iron sulphide film on the metal. Fluid velocity creates sheer stress on solid surfaces — the shear stress can be enough to rip off the imper-fect iron sulphide layer, and even a low velocity seems very likely to dislodge and remove iron carbonate deposits. When this happens, fresh iron is exposed and corrosion increases in the region of high shear stress. Subsequently, the dislodged iron sulphide and iron carbonate particles can act as scour-ing agents and increase corrosion in other areas of the plant via that mechanism. For these reasons, the gas treating industry has generally adopted velocity limits for carbon steel piping in amine service. Lean amine velocities are typically held below 7–10 ft/s while rich amine velocities are commonly held below 3–5 ft/s.

Even a little flashing can greatly increase the velocity in a pipe, and the high velocity flow from the passing biphase scours the iron sulphide layer. This results in dras-tically shorter piping life. Although less accurately than for single-phase flows, it is still possible to provide reasonable estimates of corrosion rates in two-phase flows, and this is done in the present model.

Heat stable salts (HSS) affect corrosion rates by altering solution speciation, and they chemically exacerbate corrosion by chelating iron and destroying the passivating iron sulphide layer:

FeS + 7 H

20 ↔ Fe(H

20)

6+2 + HS– + OH–

Fe(H20)

6+2 + n HCOO– ↔ + Fe(n HCOO–)(2–n) +

(6 – n)H20

There are very few quantitative measurements of the effect of HSSs

In their simplest stoichiometric forms, the basic corrosion reactions of dissolved H2S species with iron are: H

2S(aq) + Fe(s) → FeS(s) + H

2(g)

2HS-(aq) + Fe(s) → FeS(s) + H2(g) + S=(aq)

For CO2, the relevant reactions

are:

CO2(aq) + Fe(s) + H

2O → FeCO

3(s) + H

2(g)

2HCO3

-(aq) + 2Fe(s) → 2FeCO3(s) + H

2(g)

The oxidation reaction with hydrogen sulphide is faster than the reaction with bisulphide; however, the alkalinity of the amine (and ammonia) solutions means that the dissolved H2S is predominantly in the bisulphide form, with very little remaining as free molecular hydro-gen sulphide. This is also true of dissolved carbon dioxide. The concentrations of free H2S and CO2 are pH dependent and pH is a func-tion of amine strength, total dissolved acid gas, temperature, and to a lesser extent HSS concen-trations. However, heat stable salt species and their concentrations do affect the speciation of the solution, especially in lean solvents.

As discussed by Cummings et al1, the sequence of physico-chemical steps in the process of oxidising iron consists of transporting the acid from the bulk solution to the metal surface, adsorption of the acid onto the surface, reaction with iron, and transport of reaction prod-ucts back into the bulk solution. The steps are similar to what occurs in heterogeneous catalysis. The reac-tion of H2S with the iron component of various iron-based metallurgies forms solid iron sulphide and hydrogen gas, and as the reaction proceeds, the surface of the iron is changed to a mosaic of iron and sulphide ions. The surface expands by addition of sulphide, and the liberation of hydrogen gas exacer-bates the expansion. The surface layer is somewhat porous, and it adheres to the surface of the free metal. The iron-H2S and iron-HS– reactions form reaction products in completely different phases, as do

the reactions with carbon dioxide and bicarbonate. Thus, because reaction products are continually removed from the reacting solution, it follows from Le Chatelier’s princi-ple that there is a strong thermodynamic driving force powering continued corrosion.

What limits the corrosion reac-tions is primarily the amount of bare, unreacted iron that the passi-vating film leaves available at the metal surface. A secondary factor is the concentrations of dissolved reactant gases, H2S and CO2, which are other parameters. Thus, H2S, HS–, CO2 and HCO3

– all react with unprotected iron. To control the concentration of dissolved gas in their various forms, most practi-tioners adopt a rich amine acid gas loading, upper limit of <0.4 to 0.5 moles of total acid gas per mole of molecular amine.

PassivationIn a new amine unit or in one that has just been cleaned, during the period immediately following intro-duction of H2S, for example, the corrosion rate climbs exponentially. Measurements by Cummings et al.3 using a bench-scale mini-amine plant with relatively lean amine indicated in some cases over 600 mpy corrosion rate during this phase of operation. Eventually, the corrosion rate levels off and begins to fall as the iron sulphide layer increasingly occludes contact between H2S and iron (passivation). Finally, once the iron sulphide layer is established, the system settles down to a nominal (and hopefully low) residual corrosion rate. The time over which the final steady corrosion rate is attained is on the order of hours or even days. During these three phases of passivation, the solution first becomes nearly black, and then it changes to dark green, lighter green, and finally slightly amber. Chelated iron (which results from complexation) scatters light and turns otherwise contaminant-free solution to amber. The coloration depends on the size of the iron sulphide particles. The results of experiments on the time dependence of iron passivation by carbon dioxide are sparse. In one

gas ogrt.indd 2 06/03/2014 13:37

Linde AGEngineering Division, Dr.-Carl-von-Linde-Strasse 6–14, 82049 Pullach, GermanyPhone +49.89.7445-0, Fax +49.89.7445-4908, [email protected], www.linde-engineering.com

With Linde, customers rely on long time expertise in the supply of individually designed LNG plants. To deliver a high level of quality at low costs and in shorter time, Linde has translated the idea of product standardization into the LNG business, and developed the StarLNG™ standardisation concept for a capacity range between 100 and 1,500 tons per day.

The idea of StarLNG is to standardise and optimise a small-to-mid scale LNG plant based on a wide set of process variations.

This “Process Toolbox” is designed to cover about 90 percent of real-life parameters.

With this innovation, customers benefit from a scalable, flexible LNG plant design including many alternatives, based on pre-engineered documents and a 3D CAD-model for a fully modularized plant. StarLNG combines the highest safety standards, reliability, and efficiency, together with competitive CAPEX and OPEX and shortest delivery time.

StarLNG™The leading small-to-mid scale LNG plant.

gas linde.indd 1 04/03/2014 10:42

http://www.digitalrefining.com/139,sponsors,Linde_Engineering.html

on corrosion and, although data on certain HSSs are included in the current model, more data over a broader range of species and condi-tions would be very welcome.

ModelOur mechanistic, chemistry-based corrosion rate model was


developed on the basis of the sequence of steps outlined earlier, namely, diffusion of the corrosive agent to the surface, adsorption, reaction with iron, and either depo-sition or diffusion of products away from the surface. The parameters in the model have been regressed to a large number of corrosion

measurements taken in well-con-trolled bench scale flow apparatuses and static equipment using both H2S and CO2 rich amine systems and sour water. Although the details of the quantitative model are proprietary to Optimized Gas Treating, Inc., the corrosion rate in the model is a function of the important factors:

Corrosion rate = f(aH2S

,aHS

_, aCO2

, aHCO3

_, v, T)

Here: a

H2S = activity of dissolved H2S

aHS

_ = activity of bisulphide ion in solutiona

CO2 = activity of dissolved CO2

aHCO3

_ = activity of bicarbonate ion in solutionT = absolute temperaturev = superficial fluid velocity in pipe.

The model contains many of the elements that comprise the tenets discussed earlier and in a form consistent with the various steps occurring during corrosion, includ-ing adsorption isotherms and the temperature dependence of both reaction kinetics and turbulent diffusion. The model even includes corrosion by water alone as a limit-ing case in highly dilute solutions. Data for corrosion by H2S were correlated over the range of condi-tions shown in Table 1. Although scatter in the data is sizeable, if for no other reason than that repeata-ble corrosion rate measurements are hard to make, especially by weighing coupons, Figure 1 shows that this model fits the measure-ments over nearly three orders of magnitude. Data for corrosion by CO2 were taken in both a static cell and in a flow apparatus. Parameter ranges and a parity plot comparing the model with measurements of corrosion in CO2-only amine systems are shown in Table 2 and Figure 2, respectively.

In correlating the model, Arrhenius kinetics was found to fit the data well. This is in agreement with the proposal of Cummings et al.3 These authors assumed a power law kinetics expression in terms of molecular and ionic concentrations; however, we found a better fit using activities rather than concen-

100

1000

10

1Model corr

osi

on r

ate

, M

PY

0.10 1 10 100 1000

Measured corrosion rate, MPY

Rich amineSour water

Figure 1 Parity plot of measured vs model corrosion rates in rich amine and sour water systems containing H

2S only

100

1000

10

1Mod

el corr

osi

on

rate

, M

PY

0.10 1 10 100 1000

Measured corrosion rate, MPY

Figure 2 Parity plot of measured vs model corrosion rates in amines containing CO2 only

gas ogrt.indd 3 12/03/2014 12:58


3 Cummings A L, Veatch F C, Keller A E,Corrosion and corrosion control methods inaminesystemscontainingH

2S,NACECorrosion

1997Conference,PaperNo,341.

Nathan A Hatcher joined Optimized GasTreating, Buda, Texas, as Vice-President,TechnologyDevelopment,in2009.HeholdsaBSinchemicalengineeringfromtheUniversityof Kansas and is currently a member of theAmineBestPracticesGroup.Email: [email protected] E Jones joinedOptimizedGasTreating,Inc as a Software Development Engineer in2012. HeholdsaBSinchemicalengineeringfrom McNeese State University and a MS inchemical engineering from the University ofNewMexico.G Simon Weiland recently joined OptimizedGas Treating as a Software DevelopmentEngineer. He will receive a BS in chemicalengineeringfromtheUniversityofOklahomainMay2014.Ralph H Weilandisaco-founderofOptimizedGas Treating with offices in Clarita, OK,Houston, TX and Buda, TX. He holds BASc,MAScandPhDdegreesinchemicalengineeringfromtheUniversityofToronto.Email: [email protected]

underpinning lends it not only to confident interpolation, but also to reliable extrapolation to conditions well outside measured ranges.

References1 Cummings A L, Waite S W, Nelsen D K,Corrosion and corrosion enhancers in aminesystems, Brimstone Sulfur Conference, Banff,Alberta,April2005.2 PoldermanLD,DillonCP,SteelAB,Oil & Gas J.,16May1955,180.

trations. Note that neither amine strength, nor amine type, nor ammonia content appears explic-itly, although these parameters have significant effects on the activ-ity of the corrosive species in solution because speciation depends on the amine type, its concentration, and the temperature. Corrosion of other metallurgies is included in the model’s capabilities, as is the corrosion of common pipe fittings. The model’s mechanistic

Parameter RangeLabvelocity 0–25m/sTemperature 55–120°CSourwater 1–30wt%NH

4HSequivalent

3.5–10barH2Spartialpressure

Amines MEA(18–30mass%) DEA(30mass%) MDEA(45mass%) 0.1–0.8moleH

2S/moleamine

CalculatedpH 6.0–9.5

Parameter ranges for corrosion by H2S

Table 1

Parameter RangeLabvelocity 0–13.5m/sTemperature 20–160°CAmines MEA(6–30mass%) DEA(10–40mass%) AMP(9–35mass%) MDEA(35mass%) 0.0–0.4moleCO

2/moleamine

CalculatedpH 8.4–11.36

Parameter ranges for corrosion by CO2

Table 2


Ivelina Shishkova is R&D Department Manager with Lukoil Neftohim Burgas. She holds a MS in organic chemistry engineering and a PhD in petroleum refi ning from Sofi a Chemical and Technological and Metallurgical University, and has authored more than 20 technical papers. Email: [email protected] Dinkov is the Quality Manager in the Process Engineering department of Lukoil Neftohim Burgas. His research interests include crude oil characterisation, bio/conventional fuels blends characterisation and modelling of refi nery distillation processes. He holds a MS in organic chemistry engineering from Burgas University and a PhD in the technology of fossil and synthetic fuels from the University of Chemical Technology and Metallurgy, Sofi a. Email: [email protected] Jegorov is the Sales Development Manager for Grace in the CIS region. Prior to joining Grace, he was an FCC process engineer at the Mazheikiai refi nery in Lithuania. Petko Petkov is a full professor and rector of the Burgas University ‘Assen Zlatarov’. He teaches in the social science department in the fi eld of oil refi ning and lubricants, and has authored more than 180 scientifi c papers and fi ve books. Email: [email protected]

higher FCC C4 yield led to higher production of alkylate, which resulted in the production of 2% more premium grade gasoline.

References1 Watanbe K, Nagai K, Aratani N, Saka Y, Chiyoda N, Mizutani H, Techniques for octane enhancement in FCC gasoline, 20th Annual Saudi-Japan Symposium, Dhahran, December 2010. 17. Montgomery J A, Guide to Fluid Catalytic Cracking, Part 1, 1993.

Ivan Chavdarov is a Chemical Engineer in the Process Engineering department of Lukoil Neftohim Burgas, Bulgaria. His activities are focused on guiding the operation of the units of the FCC complex, troubleshooting support and optimisation of the performance of the FCC complex. Email: [email protected] Stratiev is Chief Process Engineer with Lukoil Neftohim Burgas. He holds a MS in organic chemistry engineering, and a PhD and a DSc in petroleum refi ning from the Burgas University ‘Assen Zlatarov’. He has authored more than 130 papers. Email: [email protected]

Regular A-92

Premium A-95

Super A-98

RVP = 60kPa

1%28%

71%

RVP = 50kPa

1%26%

73%

Figure 3 Effect of changing the RVP on refi nery gasoline grades produced during the Resolution catalyst period

MINIDIS ADXpert

Grabner Instruments Messtechnik GmbH | A-1220 Vienna, Austria | Dr. Otto-Neurath-Gasse 1Phone +43-1-282 16 27-0 | Fax +43-1-280 73 34 | [email protected]

QUICK, ACCURATE, PORTABLE

DISTILLATION

Portable, True Atmospheric D7344 Distillation•Excellent D86, ISO 3405, IP 123 Correlation•Fully Automated Measuring and Cleaning•6 mL of Sample•No Glassware, no Open Flame•Distillation of Gasoline, Jet Fuel, Diesel, Solvents, •Unknown Samples, Raffinates and BiodieselSpecified for Use by US Pipelines•

Grabner Instruments Messtechnik GmbH | A-1220 Vienna, Austria | Dr. Otto-Neurath-Gasse 1

Portable, True Atmospheric D7344 DistillationExcellent D86, ISO 3405, IP 123 CorrelationFully Automated Measuring and Cleaning

Distillation of Gasoline, Jet Fuel, Diesel, Solvents, Unknown Samples, Raffinates and Biodiesel

ADXpert half page 1.indd 1 06.12.2013 11:46:37

grace.indd 4 11/12/2013 12:49gas ogrt.indd 4 11/03/2014 12:32

http://www.digitalrefining.com/153,sponsors,Grabner_Instruments.html

enersul.indd 1 10/03/2014 11:01

http://www.digitalrefining.com/164,sponsors,Enersul_LP.html

Zeolite based dryers in ethylene plants

Steam cracking represents more than 95% of ethylene produc-tion. This article focuses on

product stream drying duties in ethylene production based on steam cracking of paraffinic hydrocarbons.

Steam cracker feeds can be sepa-rated into two categories:• Natural gas liquids (NGLs): ethane, propane and butane• Heavier hydrocarbons such as naphtha and gasoils.

North America (>5%) and the

A review of the role and nature of adsorbent dryers in the production of ethylene by steam cracking

VASSILIOS ZAFIRAKIS and HANS HOEFERGrace Materials Technologies

Middle East (>80%) use gaseous feedstock to produce ethylene, whereas in Europe (>80%), in Asia-Pacific (>81%) and in Latin America (>65%) naphtha and gasoil feeds dominate. On the world scale, ethylene feedstocks are as follows:• Naphtha 52%• Gasoil 6%• Ethane 28%• Propane 8%• Butane 3%• Other 3%

Process sections of an ethyleneplantIn general, ethylene production has four major sections:1. Cracking heater section including transfer line heat exchangers (TLE)2. Gasoline fractionator and quench water tower 3. Compression section with both acid gas removal and a main drying unit4. Chilling train and separation


Compression

Compression

Methanising

Deethanising

Depropanising

C2H2 hydrogenation

C3H4 hydrogenation

Demethanising

Acid gas removal

Drying

PSA

Cracking

C2-cutdrying

C3-cutdrying

H2 drying

Water quench

Chilling

Ethylene fractionation

Propene fractionation

Feedstock liquid

High purity hydrogen

Medium purity hydrogen

Methane

Ethylene

Propene

TLE and oil quenching

Primary fractionation

Pyrolysis fuel oil

Debutanising

C4-cutPyrolysis gasoline

Figure 1 Simplified process flow diagram for ethylene production by liquid cracking with a front-end demethaniser. In green boxes: zeolite based adsorbers.

gas grace.indd 1 11/03/2014 09:30


550-650°C. The lowest temperature refers to the start of the cracking cycle (SOR), while the highest temperature represents the maxi-mum temperature, where the coil and the TLE need to be cleaned (EOR) and the heater has to be shut down.

Gasoline fractionator and quenchwater towerOne of the major differences between liquid and gas crackers is the use of a gasoline fractionator. Liquid crackers require gasoline or primary fractionators. This is not necessary for crackers using ethane or propane as feedstock (see Figure 2).

Furthermore:• For liquid feedstock downstream of the TLEs, the required cooling to temperatures of 200-230°C cannot be performed in standard shell and tube heat exchangers; condensation of a significant portion of liquids in the cracked gas leads to excessive fouling by carbon deposits.

section for cracked gas separation into the desired products.

A variety of commercial process routes is available. Whereas crack-ing heaters, quench towers and compression sections are similarly configured (sections 1-3 above), for separation (section 4) three main process routes are established. They are characterised by the first sepa-ration tower and the position of the unit for hydrogenation of C2 acety-lenes in the cracked gas. These three process sequences are:1. Demethaniser first with tail-end hydrogenation2. Deethaniser first with front end hydrogenation 3. Depropaniser first with front end hydrogenation.

Cracking heater section and TLEsThe ethylene industry is served by a limited number of licensors and contractors with their own ethylene process technologies. They have designed and built most of the world’s ethylene plants. In the

process route shown in Figures 1, 2, 3 and 4 hydrocarbon feedstock (ethane, propane, butane, naphtha and gasoil) is fed into a pyrolysis heater (furnace) and preheated to temperatures of 500-650°C. Steam enters the heater coil and the hydrocarbon/steam mixture is further heated to temperatures of 750-890°C. Within this range, satu-rated hydrocarbon molecules crack to form olefins such as ethylene, propylene and butene, and diole-fins, as well as other hydrocarbon molecules such as methane, benzene, toluene, and hydrogen and other products. After this pyrolysis reaction the produced gas from the heater coils is rapidly cooled in tubular heat exchangers (transfer line exchangers, TLE) against boiler feed water, produc-ing high pressure steam (HPS). In the case of gaseous feeds, the gas leaves the TLE at a temperature of 300-400°C. For liquid feeds such as light to heavy naphtha, tempera-tures are 420-450°C, and for gasoil

Compression

Compression

Methanising

Deethanising

Depropanising

C2H2 hydrogenation

C3H4 hydrogenation

Demethanising

TLE Acid gas removal

Drying

Drying

PSA

Cracking

C2-cutdrying

C3-cutdrying

H2 drying

Water quench

Chilling



Feedstock gas

Gasoline



Methane

Ethylene

Propene

C4+ hydrocarbons

Figure 2 Simplified process flow diagram for ethylene production by gas cracking with a front-end demethaniser. In green boxes: zeolite based adsorbers

gas grace.indd 2 11/03/2014 09:43

Therefore, the cracked gas is further cooled by injection of oil (oil quench) by means of devices called quench fittings. Oil quenching is performed downstream of the TLEs after each heater, or in the combined cracked gas line (transfer line) from all heaters. The mixed two-phase flow of cracked gas and quench oil is separated in the bottom section of the primary frac-tionator. In this tower the cracked gas is separated from the fuel oil, which leaves the tower from the bottom. The overhead is directed to the quench water tower• For gaseous feedstock, cracked gas from the TLEs is further cooled to about 200°C against feed or boiler feed water (BFW) in the so-called secondary TLE, before entering the quench water tower where the cracked gas is further cooled.

In a liquids cracking plant, this tower essentially works as a partial condenser to the upstream oil frac-tionators, which condense both the


steam and a significant part of the gasoline components. Part of the gasoline is used as reflux for the upstream oil fractionator. The remaining gasoline leaves the unit after stripping in a side tower for stabilisation.

The cracked gas leaves the top of the quench tower at a temperature of 35-40°C, slightly above atmos-pheric pressure and enters the compression section of the plant.

Compressor section, acid gas removal and drying of cracked gasThe cracked gas is compressed in four to five stages to pressures of 35-37 bar. Usually, compression of the charge gas from gaseous feed-stock cracking takes place in centrifugal compressors employing four stages, while five compressor stages are required for compression of cracked gas from the cracking of liquid feedstocks such as naphtha or gasoil. Consequently, in plants processing gaseous feedstock a higher compression ratio is

employed, yielding a higher discharge temperature compared to liquid based plants. The number of stages depends primarily on gas composition and the highest temperature allowed for the discharge of the different compres-sion stages. Such a temperature is typically below 100°C in order to avoid excessive equipment fouling through polymerisation of diolefins and other precursors in the cracked gas. As the pressure increases at the different compressor stages, water and hydrocarbons are condensed in the downstream cool-ers and separated from the gas in the interstage separators.

Water is returned to the quench water tower while the hydrocar-bons from the first three stages are recycled to the gasoline stripper in the upstream hot section of the plant. The acid gas removal system is usually positioned after the third or fourth compression stage. Cracked gas leaving the acid gas removal unit has a remaining acid

Compression

Compression

Methanising

Deethanising

Debutanising Depropanising

C2H2 hydrogenation

C3H4 hydrogenation

Chilling and demethanising


Acid gas removal


Drying

Drying

PSA

Cracking

C2-cutdrying

C3-cutdrying

H2 drying

Water quench

Chilling



Feedstock liquid

Pyrolysis fuel oil



Methane

Ethylene

Propene


Figure 3 Simplified process flow diagram for ethylene production by liquid cracking with a front-end deethaniser. In green boxes: zeolite based adsorbers

gas grace.indd 3 11/03/2014 09:44


purity is withdrawn from the lowest temperature separator; this can be further purified in a down-stream system. The condensates produced over several stages are fed to the demethaniser (see Figure 1). In recent plants, mixed refriger-ants are used instead of cascade refrigeration systems.

The process diagrams in Figures 3 and 4 show simplified process routes for liquid cracking plants using a separation sequence of deethaniser first (see Figure 3) or depropaniser first (see Figure 4).2,5

The separation section The layout of the separation section depends on the selected separation sequence, the feedstock and the targeted products. The following represents the demethaniser first concept (see Figure 1):• Demethaniser• Deethaniser, acetylene converter, acetylene converter dryer• Ethylene fractionator (splitter)

ment. Therefore the condensate always contains traces of free water. This needs to be considered in the design of downstream condensate dryers. Complete removal of water is mandatory for the charged gas or liquid streams to be cooled to temperatures below 15°C in the downstream chilling train and the separation section, to avoid the plugging of lines and equipment due to formation of hydrates and ice. For the removal of moisture from these streams, the most current technology uses solid alumino-silicates (zeolitic molecular sieves) which will be discussed later. Previously, dual systems (glycol followed by alumina) were employed.

After drying, the cracked gas is partially condensed in several stages of a cascade refrigeration system (propylene, ethylene, meth-ane), resulting in cold product streams at temperatures of -135°C to -145°C. Hydrogen of 95% mole

gas content of about 1 ppmv. Depending on the technology selected downstream of the acid gas removal system, one or two additional compression stages may be employed. Additional C2- and C3 rich condensate is produced in these compression stages. This is fed into the stripper, where the C2 and lighter fractions are forced through the tower top back to a compression section operating at lower pressure. The bottoms of this tower are sent to the downstream depropaniser. If the operating temperature in the depropaniser is lower than 15°C, the bottom stream has to be dried to avoid the forma-tion of hydrates.

The cracked gas is saturated with water before compression as well as after each intercooler stage and down-stream of the acid gas removal system. The condensate is also saturated with water, while excess water separation depends on the design of the separator equip-

Compression

Compression

Methanising

DeethanisingDepropaniser C2H2 / C3H4 hydrogenation

C3H4 hydrogenation

Chilling and demethanising

Acid gas removal

PSA

Cracking

C2-cutdrying

C3-cutdrying

H2 drying

Water quench



Feedstock liquid



Methane

Ethylene

Propene



Pyrolysis fuel oil

Debutanising


Drying

Chilling

Figure 4 Simplified process flow diagram for ethylene production by liquid cracking with a front-end depropaniser. In green boxes: zeolite based adsorbers

gas grace.indd 4 11/03/2014 09:44


provider may minimise by chemi-cal modifi cation of the selected adsorbent.

Kinetics or, better, mass transfer kinetics, is the overall term related to intra-particle mass transfer char-acteristics causing breakthrough behaviour as well as lost bed mass. Lost bed mass is that part of the adsorber bed which is still active even though the breakthrough has just started. Fast kinetics provides a sharp breakthrough curve with a small lost bed mass, while slow

to be adsorbed (the adsorbate). In practice, one has to distinguish between equilibrium capacity, given through isotherms, and the typically lower overall dynamic capacity, which is characterised by specifi c operating conditions.

Selectivity is the ratio of the amount of purposely adsorbed component to that of all others in a fl uid to be processed. In reality, selectivity is never 100% for the targeted adsorbate due to co-adsorption, which the adsorbent

• Depropaniser, methyl acetylene and propadiene (MAPD) converter, MAPD converter dryers• Propylene fractionator (splitter)• Debutaniser.

The purpose of the demethaniser is the separation of methane and other lights (CO, H2) from C2+ components which are sent to the deethaniser. This tower separates C2 components (acetylene, ethylene, ethane) from heavier hydrocarbons. C2 hydrocarbons leave the deetha-niser as the overhead product, while the C3 and heavier compo-nents, the bottoms, are directed to the depropaniser. The overhead of the deethaniser is heated and hydrogenated in order to convert the acetylenes into ethylene and ethane. Because traces of water may be present during this hydro-genation, drying of the C2 hydrocarbons is necessary before they enter the ethylene fractionator, typically operating at -10°C to -35°C and 16-18 bar. The ethylene product is the overhead of the ethylene fractionator. The bottom product of the ethylene fractionator is ethane, which is recycled back to the heaters for cracking to extinc-tion. In the depropaniser, C3+ hydrocarbons are separated into two fractions. The C3 fraction as overhead and the heavier C4 frac-tion leaving the tower as bottom product are sent to the debutaniser. Due to the very low C3+ fraction in ethane based crackers, C3 and C4 separators are not present.

Zeolitic adsorbents in ethylene plantsAny adsorbent has basically seven features, six which qualify such material for specifi c adsorption duties, namely: • Capacity• Selectivity• Kinetics • Process compatibility • Regenerability • Mechanical stability (crush strength, attrition)• Cost.

Adsorption capacity is the amount of adsorbate related to adsorbent mass, which is a function of operating temperature and the partial pressure of the component

SweetSulf TM

Sprex®

AdvaSulf TM

AdvAmine TM

COSWEET TM

HySWEET®

A UNIQUE TASTE OF SWEET FOR YOUR GAS

50 YEARS OF EXPERIENCE IN GAS SWEETENING AND SULPHUR RECOVERY PROCESSES

With its unique and complete proprietary technologies portfolio, PROSERNAT offers optimized solutions to bring on specs any type of gas contaminated with CO2, H2S, COS and organic sulphur species, while producing sulphur with the most stringent emissions standards.

www.prosernat.com

gas grace.indd 5 11/03/2014 12:33

http://www.eptq.com/jump.aspx?a=Prosernat&p=gas14


increasing load, and measuring the required crushing force. The results vary greatly, depending on the samples’ geometries and shapes. Comparative measurements with this method are meaningful only for samples with the same geomet-rical shape and the same size distribution for beads, and the same D and L/D ratio for extru-dates. Direct comparison of beads and extrudates is less useful due to different types of force applied to the samples, for instance point vs linear or even area-shaped. A more realistic comparison divides the measured force by the length of the extrudate (N/mm), as recom-mended by ASTM.

Bulk crush strength is commonly measured by placing a representa-tive sample of beads or extrudates in a defined cylinder, applying a force with a piston, and measuring the amount of fines through a 420µ mesh. Bulk crush strength is expressed as the pressure (MPa = N/mm2) applied to generate 1% fines. This method best represents the real conditions in a fixed bed adsorber and is increasingly used.

Attrition and abrasion tests give information about the expected behaviour of an adsorbent during processes such as packaging, trans-portation, loading, as well as during operation in the fixed bed. Bulk crush strength together with attri-tion tests are of great importance. They simulate dust formation by attrition due to mechanical parti-cle-particle interaction in a fixed bed. Adsorbents with high bulk crush strength and low attrition and abrasion can be produced by apply-ing sophisticated beading or pelletising procedures during the adsorbent manufacturing process. When comparing materials from different suppliers, it is important to ensure that the measuring method is the same, otherwise the results are misleading.

The cost of the adsorbent comprises its purchase and unit operating costs. Selectivity, kinetics, compatibility and regenerability directly impact the performance and lifetime summarised in the productivity and financial feasibil-ity of any gas purification process.

with a boiling point range of 120-400°C. The adsorbent provider has to reduce the concentration of such active sites, or deactivate them by masking with chemical additives.

Regenerability ensures adsorption and desorption in sequential cycles with uniform performance. The adsorbate must be reversibly physisorbed. Irreversible side reac-tions and the accumulation of such reaction products in the pore system of the zeolite, as well as the deterioration of the working bed capacity, are often described as adsorber bed ageing. Reducing working bed capacity goes with increased operating costs due to shorter cycle times, higher amounts of regeneration gas and increased pressure drop, finally resulting in a changeout.

Another issue is the mechanical stability of the adsorbent in the fixed bed. Commercial adsorbents need to have adequate mechanical strength so that losses during handling and processing are mini-mised. Three measured parameters — single piece crush strength (ASTM D-4179, ASTM D-6175), bulk crush strength (ASTM 7084-04) and attrition/abrasion (ASTM 4058) — describe expected mechan-ical behaviour. Single piece crush strength is measured by placing one adsorbent body, bead or extru-date between two plates, applying

kinetics leads to an extended curve and a large lost bed mass. The effect of an extended breakthrough curve can be counteracted either by using more adsorbent or by accept-ing reduced throughput per given adsorption cycle. A third option is the use of smaller adsorbent bodies offering a larger overall adsorbent bed surface. As a rule of thumb, one can say that the mass transfer zone of 1.6 mm to 3 mm beads is just a third that of 2.5 mm to 5 mm beads. The smaller beads, however, lead to a higher pressure drop. Pressure drop and mass transfer zone length can be optimised by using a split bed. In typical bed designs, smaller beads make up the mass transfer zone while the larger beads form the equilibrium zone.

Process compatibility describes a number of effects during the adsorption/desorption process which may reduce the lifetime of the adsorber bed. In an ethylene plant, the charge gas contains very reactive components such as acety-lenes, olefins, diolefins and styrene. They lead to oligomerisation/polymerisation and finally to coking during regeneration. The carbon rich mass from heavier hydrocarbons formed at catalyti-cally active sites of the desiccant body is also known as ‘green oil’. This represents the most critical compatibility issue. Green oil is a mixture of mainly C4-20 oligomers

4 A type Zeolite 3 A type Zeolite

Atoms not shown to scale

Ion exchangeNa+

Na+

Na+

Na+

Na+

K+

K+

K+

AluminiumSiliconOxygen

CationsNa+

K+

Figure 5 Schematic representation of the Na+/K+ ion exchange of an A-type zeolite struc-ture with one sodalith cage and eight truncated octahedra at its corners

gas grace.indd 6 11/03/2014 17:09

MAN Diesel & Turbo has unique compression solutions in its portfolio for the complete range of CO2, N2, propylene and vapour related applications, with single machines or complete train solutions. Others talk about CO2 compression. We have the credentials! Over 200 RG Gas compressor units are in operation or on order. We are world`s number one in CO2 high pressure applications, thanks to sophisticated test-ing facilities and proven track records in the fi eld. With an RG Gas compressor from MAN Diesel & Turbo you will gain an exceptionally fl exible and optimized solution to maintain your business economic.Find out more at www.mandieselturbo.com

Marine Engines & Systems Power Plants Turbomachinery After Sales

Actions speak louder than wordsHigh pressure CO2 gas compression

Typ 02 - Actions speak louder than words (RG Gas) - 297x210.indd 1 2014-01-27 11:59:29man.indd 1 04/03/2014 10:47

http://www.eptq.com/jump.aspx?a=MAN&p=gas14

FLEXSORB® —Leader in H2S Removal

Sour Gas Treating Problems Solved ✓ Absorbs the H2S, Rejects the CO2

✓ Cost-Effective for Grassroots and Retrofits

✓ Removes H2S to Less Than 10 ppm

✓ Vast Commercial Experience (100 + units)

✓ Enables Simple, Low Cost Retrofits

3225 Gallows Road, Fairfax, Virginia 22037-0001, USAwww.exxonmobil.com/refi ningtechnologies • +1-703-846-2568 • fax +1-703-846-3872 • [email protected]

exxon.indd 1 07/03/2014 09:52

http://www.digitalrefining.com/74,sponsors,ExxonMobil_Research_and_Engineering.html


molecules are positioned during the adsorbing phase, relying on strong physical forces called physisorp-tion; this is fully reversible, in contrast to chemisorption where chemical reactions are involved.

plants, because it reduces product losses through low co-adsorption.

The 3A zeolite has a channel system with precisely dimensioned openings and negatively charged cages. In these cages, adsorbed

The cost however, which may be generated by a production loss due to a malfunction of the adsorption unit, is far higher than the cost of a changeout of the unit. Reported cases comprise the selection of non-process-compatible adsorbents and/or a design which is not adequate for the process unit. Therefore, during design and changeout of an adsorber unit, consultation with experienced adsorbent suppliers is inevitable for ensuring reliable unit operation.

Industrially relevant adsorber zeolites are artificially crystallised alkaline and alkaline earth alumi-no-silicates, which have the general formula (M’, M’’)•[(AlO2)x•(SiO2)y] •zH2O. Such zeolites are crystalline solids consisting of tetrahedral SiO4 and AlO4 building blocks, forming microporous structures. Each aluminum tetrahedron carries a negative charge, which has to be neutralised with a metal ion (M’, M’’) such as Na+, K+ and/or Ca2+. Per unit cell, the SiO4 and AlO4 tetrahedra form eight truncated octahedra, which build one ‘soda-lith cage’, and are stacked in a cube resulting in an A-type zeolite.

The 3A type zeolite, the work-horse of ethylene plant dryers, is produced via Na+ -> K+ ion exchange, starting from an 4A type zeolite with a pore opening of about 4Å (0.4 nm), whereas the larger K+ ions narrow the pore openings by about 1Å (0.1 nm, see Figure 5).

The impact of Na+/K+ ion exchange on the dimension of A-type zeolite pore orifices, and the subsequent change of equilibrium adsorption capacities for various adsorbates, is shown in Figure 6. The adsorption of all adsorbates is reduced when K+ concentration increases. Reduction of water capacity, however, is much less pronounced than for any other adsorbate shown. In ethylene drying, an A-type zeolite with a K+

exchange of 50% already shows efficient moisture pick-up features and, in parallel, low adsorption capability for ethylene. This is important for the application of 3A zeolite in drying ethylene rich streams, as occurs in ethylene

Carbon dioxide at 700 torr, 25ºC

Ethylene at 700 torr, 25ºCEthane at 700 torr, 25ºCMethanol at 4.0 torr, 25ºC

Water at 4.5 torr, 25ºC

Oxygen at 700 torr, –183ºC

0.15

0.25

0.30

0.20

0.10

0.05

gab

sorb

ate /g

abso

rbent

00 0.1 0.2 0.3 0.4 50. 0.6 0.7 0.8 0.9 1.0

Exchange of K, K+/ (K++Na+)

Figure 6 Impact of Na+/K+ ion exchange on equilibrium adsorption capacities measured on powders7

Figure 7a Ball-and-stick representation of the structure Figure 7b Microporous channels with openings of a 3A zeolite, consisting of eight sodalith cages. 3Å for moisture transportation Figure 7c Sodalith cages with diameter >0.8 nm Figure 7d 3A zeolite pore structure comprising moisture adsorbance channels and cages for moisture transportation and storage

a b

c d

gas grace.indd 7 11/03/2014 09:44


tions are 15-20 wt%. This reduces the nominal equilibrium adsorption capacity by approximately the same percentage because clay based binders have limited to no adsorp-tion properties.

The binder is fundamentally important to both the mechanical stability and the catalytic activity of the molecular sieve body. Figure 9 shows the principal dependence of the amount of used binder on the crush strength of the zeolite bead and its corresponding water equi-librium capacity. With increasing binder concentration the crush strength also increases but, at the same time, the equilibrium water adsorption capacity deteriorates.

Catalytic activity in the adsorbent body has to be minimised. Consequently, for ethylene plant adsorbents reactive metal-free clay binders have to be selected.

Beside the degree of K+ exchange and the type and amount of binder, the firing conditions for beads in the adsorbent manufacturer’s calciner play an important role in the bead’s performance, in terms of mechanical stability and adsorption properties. The two key process parameters are calcination tempera-ture and residence time. These parameters also impact co-adsorp-tion behaviour towards ethylene and eventually to polymerisation and coking.

In summary, the selected optimal binder ensures a number of impor-tant characteristics for the final adsorber bead, namely:• The right meso- and macroscopic transportation pore configuration • A mechanically stable bead with low attrition • Minimised catalytic activity when exposed to the reactive species of hydrocarbon feedstock.

The uniform crystalline structure of zeolitic molecular sieves, in combination with a secondary transportation pore system, provides predictable and reliable adsorption properties.

Typically, zeolitic adsorber beads are commercially available in bead sizes ranging from 1 mm to 5mm in diameter. Depending on the adsor-bent duty, bead size is selected by trading adsorption kinetics with

molecules with the right dynamic dimensions may enter the pores, whereas larger molecules will be rejected. Figure 8 depicts a symbolic 3A zeolite structure with water molecules adsorbed from a wet ethylene stream.

After crystallisation resulting in a particle size of approximately 10µ, the individual A-type zeolite crys-tals, however, are far too small for use in fixed bed adsorbers. Consequently, they are mixed with a clay binder for moulding into macroscopic beads. Around binder zeolite contact points the pores of the zeolite will be at least partly blocked. Typically binder propor-

The interconnecting channels allow the adsorbate molecules to travel into the cages (adsorption) or leave them (desorption). Both channels and cages build up the 3A zeolite pore system shown in Figures 7a to 7d, according to the University of Princeton’s ZEOMICS library.8

The red spheres represent oxygen atoms, the sticks are oxygen-alumi-num and oxygen-silicon bonds. For a better view of the Na+ ions, the K+ ions and the truncated octahedra are omitted.

The 3A zeolite framework of symmetrically orientated tetrahedra and octahedra and its pore struc-ture is stereo-selective; smaller

15

25

20

10

5

Wate

r adso

rpti

on c

ap

acit

y,

wt%

00 5 10 15 20 25 30

Amount of binder, wt%

Water adsorption capacity Crush strength (arbitrary units)

Figure 9 Principal dependence of binder amount on crush strength and equilibrium water capacity9

Figure 8 Idealised stereo-selective moisture adsorption on a 3A zeolite

gas grace.indd 8 11/03/2014 09:44

Lewis® Pumps comprehensive product line includes:

Sulphuric Acid Pumps

Molten Sulphur Pumps

Axial Flow Phosphoric Acid Pumps

Molten Salt Pumps

Sulphuric Acid Valves

• Butterfly Valves

• Gate Valves

• Globe Valves

Lewmet® Specialty Items

• Dilution Pipes

• Orifice Plates

• Strainers

• Thimbles

• Vortex Breakers

To learn more about Lewis® Pumps or Weir Minerals products, visit: www.weirminerals.com/Lewis

Weir Minerals Lewis Pumps is known internationally in the sulphur, sulphuric acid and phosphoric acid industries with equipment installed in more than 120 countries worldwide. Founded in 1891 as Chas. S. Lewis & Co., Inc., today Weir Minerals Lewis Pumps is the recognized world leader for pumps and valves in these difficult pumping applications.

LEWIS® PUMPS Vertical Chemical Pumps

The world standard for vertical pumps and services to the sulphur chemicals industry

Weir Minerals Lewis Pumps8625 Grant Rd.St. Louis, MO 63123

T: +1 314 843-4437F: +1 314 843-7964Email: [email protected]

Copyright © 2013, EnviroTech Pumpsystems, Inc. All rights reserved. LEWMET, LEWIS, and LEWIS PUMPS are registered trademarks of Envirotech Pumpsystems, Inc.; WEIR is a registered trademark of Weir Engineering Services Ltd.

weir .indd 1 06/03/2014 12:06

http://www.eptq.com/jump.aspx?a=Weir&p=gas14

AMACS 24-7 Emergency Service Available at (281) 716 - 1179 © 2013 Amacs Process Towers Internals. All Rights Reserved. Member of Fractionation Research

www.amacs.com

CARTRIDGE TRAYS

DUAL FLOW GALLERYDISTRIBUTOR

VALVE BUBBLE CAP

TRAYS FOR EVERY PROCESS

For gas processing, refining and a wide rangeof process applications. AMACS offers a wideselection of alloys, sizes and configurationsto meet any mass transfer requirement. As aleader in the research and design of trays andtower internals, AMACS can design to yourspecifi cations or improve your process with our latest technologies.

• Valve (numerous options)

• Sieve or perforated

• Bubble cap

• Cartridge trays

• Dual Flow

• Gallery Distributor Tray

For all your process internal requirements make one stop- AMACS.

SIEVE OR PERFORATED


For gas processing, refining and a wide rangof process applications. AMACS offerofferoff s a widselection of alloys, sizes and configurationto meet any mass transfersfersf requirement. Asleader in the research and design of traystower internals, AMACS can design to youspecifi catfi catfi ions or improve your process with our latest technologies.



• Bubble cap

• Cartridge trays

• Dual Flow




For gas processing, refi ning and a wide range of process applications. AMACS offers a wide selection of alloys, sizes and confi gurations to meet any mass transfer requirement. As a leader in the research and design of trays and tower internals, AMACS can design to your specifi cations or improve your process with our latest technologies.



• Bubble cap

• Cartridge trays

• Dual Flow



AMACS 24-7 Emergency Service Available at (281) 716 - 1179 © 2013 Amacs Process Towers Internals. All Rights Reserved.

www.amacs.com

amacs gas.indd 1 11/03/2014 09:18

http://www.digitalrefining.com/56,sponsors,AMACS_Process_Tower_Internals.html


for adsorbate molecules is the driv-ing force. In depressurisation mode, the pressure inside the adsorber vessel will be lowered by a factor > 10. At low pressures the affinity of the zeolite pores for the adsorbate molecules is significantly reduced, resulting in adsorbate release from the adsorber bed. When the sorbent material has been sufficiently freed of adsorbate, the adsorber is ready for the next cycle.

In order to allow continuous TSA/PSA gas conditioning, two or more adsorber beds in the gas treatment unit are in adsorption, regeneration or standby modes.

Dryers in ethylene plantsComplete removal of the water contained in cracked gas is manda-tory in order to avoid plugging of downstream cryogenic sections due to freezing or formation of hydrates. Depending on the cracked feedstock and the configu-ration of the ethylene plant the following dryers may be in opera-tion (the first three are mandatory for all ethylene plants):1. The charge gas dryer or main ethylene dryer, located after the acid gas removal unit2. The condensate dryer for condi-tioning the gasoline stream from the condensate stripper3. The acetylene converter or secondary dryer downstream of the C2 acetylene converter4. The MAPD converter dryer downstream of the MAPD converter

its accepted maximum impurities level, the adsorber has to be regenerated.

Two regeneration methods are used: thermo-swing adsorption/desorption (TSA) and pressure swing adsorption/desorption (PSA). They rely on the fact that, apart from the physic-chemical nature of the adsorbate, adsorption and desorption are pressure, temperature and time controlled. In TSA units, temperature is the process variable; in PSA units pres-sure is the variable.

In an ethylene plant, with the exception of the Hydrogen PSA purifier, all other adsorption units, including the hydrogen dryer downstream of the methanator, are TSA types. In the TSA case, thermal regeneration is accomplished by passing hot product gas through the bed, heating it up to a tempera-ture significantly above the temperature during adsorption. The hot adsorber bed decreases its affinity for adsorbates and releases them into the gas flow now acting as purge fluid. After completion, the adsorber bed is cooled down, regaining its adsorption capacity. Then the unit is ready for the next cycle. Typical regeneration gas temperatures are 240°C for 3A molecular sieves and 250-280°C for 4A molecular sieves.

In contrast, pressure swing regen-eration takes place at the same temperature as adsorption. In the adsorption mode at elevated pres-sures the affinity of zeolite pores

pressure drop and mechanical stability. As a rough guide, the smaller the beads the better the kinetics, but the higher the pressure drop the lower the single piece crush strength, and vice versa. Commercial adsorber beads have a comparatively narrow bead size range in order to have as close as possible to mono-disperse particle distribution characteristics targeting adsorber bed porosities of 30% to 35%. Depending on the supplier´s philosophy, zeolitic adsorber extru-dates or pellets with L/D ratios from nearly 1 to 5 are in use. Extrudates with higher L/D ratios may lead to a lower space filling and a lower bulk density. They tend to have favourable mass trans-fer properties, however, sacrificing adsorption capacity and jeopardis-ing the adsorber bed’s mechanical stability. The pressure drop for beads is more favourable than for extrudates (see Figure 10), even in the typical gas velocity range between 5 to 12 m/min.

Even more importantly, plant operators occasionally report signif-icantly increased dust formation when loading and unloading the adsorber vessel, as a result of extru-date breakage and split-offs.

Adsorber zeolites have a strong affinity for polar and polarisable molecules, whereas the higher the degree of polarity, the higher the selective adsorption power. Since water is the most polar molecule known, it is the preferred adsorb-ate. In a moisture-containing hydrocarbon stream, next to water heavier hydrocarbons will be picked up first.

Depending on the water/hydro-carbon ratio, for some time the zeolite adsorber bed will simultane-ously pick up water and some hydrocarbons. At the end of the adsorption cycle, the part of the adsorber bed next to the inlet is enriched with adsorbed water, whereas the middle part of the bed contains a higher concentration of hydrocarbons. The outlet part of the bed carries the lightest hydro-carbons. This effect is given by the higher polarity of water over heav-ier and lighter hydrocarbons. Before the product gas has reached

12

20

18

16

14

10

8

6

4

2

Pre

ssure

dro

p,

mbar/

m

00 2 4 6 8 10 12 14 16 18 20

Superficial velocity, m/min

2.5-5.0 mm beads 3.7 mm extrudates

Figure 10 Pressure drop development of beads vs extrudates

gas grace.indd 9 11/03/2014 09:44


stage at lower operating pressure (see Figure 3). This is often used when the ‘depropaniser fi rst’ with front end hydrogenation route has been chosen.

The compressor section of an ethylene plant is prone to fouling, given the nature of the cracked gas, which may contain varying concen-trations of hydrogen, paraffi ns and aromatics as well as reactive compounds such as acetylenes, olefi ns and diolefi ns. Under certain conditions, compounds such as butadiene, styrene, isoprene and vinyl acetylene may form polymers. Consequently, charged gas dryers, which are embedded in this section, are also impacted by this fouling tendency. Such fouling is the result of oligomerisation, polymerisation and condensation reactions, given by:• Free radical polymerisation reactions• Diels-Alder polymerisation of active Dienes (Cyclopentadiene, Methyl cyclopentadiene, and so on).

The radical polymerisation reac-tions are caused by heat alone, but

hydrocarbons into the regeneration off-gas due to absorption and co-adsorption outperforming the dual system of glycol based absorb-ers followed by alumina adsorbers• More economical to regenerate.

Two design options with differ-ent operating pressures are used, depending on the selected process route, which also defi ne the posi-tions of the dryers in the ethylene plant:1. Dryers operated at high pressure (26-37 bar, 13-15°C)2. Dryers operated at low pressure (12-25 bar, 12-14°C).

The dryer operating temperature is usually set at 2-3°C above the hydrate formation temperature. It has to be kept in mind that a temperature rise by 1°C results in a 6.5% increase in water load to be removed by the dryer. Therefore careful temperature control is advised.

Typically, the dryers are posi-tioned after the last compression stage, as in option 1 (see Figures 1 and 2). In option 2, the dryers are positioned after the next-to-last

5. The hydrogen dryer downstream of the methanator for upgrading the hydrogen stream from the cold box and or the PSA unit for hydro-gen of high purity.

The charge gas dryers or maincracked gas dryersThe main cracked gas dryer is posi-tioned in the compressor section downstream of the acid removal unit (caustic and water wash tower). Complete removal of water to a level below <1 ppmv is manda-tory. A specially developed 3A molecular sieve is state-of-the-art technology for operating such dryers. It has almost completely replaced the dual system of glycol absorbers followed by alumina adsorbers. The main technical and commercial reasons for replacing earlier drying concepts with zeolitic molecular sieves are: • Achieving gas specifi cations <0.1 ppmv water content• Achieving signifi cantly lower ‘green oil’ formation • Achieving longer on-stream time • Signifi cantly lower losses of

We make technology work.

www.hoerbiger.com

Like musicians set the tone, HOERBIGER sets standards in the Oil and Gas Industry.

Image_ad_KT_2014.indd 1 06.03.2014 15:46:22

gas grace.indd 10 11/03/2014 09:45

http://www.digitalrefining.com/151,sponsors,Hoerbiger_Compression_Technology.html


shows a typical aging curve and the corresponding pressure drop development of a naphtha based ethylene plant’s charged gas dryer.

are enhanced by peroxides. Also, the Diels-Alder condensation reac-tion contributes to fouling due to formation of heavy hydrocarbons condensing on the metal surfaces and gradually dehydrogenating. Both polymerisation reactions are catalysed in the presence of iron. The temperature is a key variable for such reactions.10,11

Grace and others encountered the following experiences regarding the operation of charge gas dryers: •All gas cracking units show a high fouling rate in the compressor section, whereas plants based on liquid cracking report low to moderate fouling. •Lower pressure charge gas dryers have to cope with a higher water load (see Table 1) and are more prone to fouling compared to those plants with the dryers set up at higher pressures (32-37 bar vs 12-22 bar) after the last compression section. This fouling is caused by non-condensed and non-separated heavier fouling precursors. Regarding the first experience, it

is known that gas-based plants, typically operating four compressor stages, generate higher discharge gas temperatures because of the higher compression ratios per stage. This effect is further enhanced by higher concentration of hydrogen and its higher isotropic exponent k=Cp/Cv:

Ts, PS = Suction temperature, pressure TD, PD = Discharge temperature, pressure

By contrast, higher content of aromatics in the cracked gas from liquid based crackers helps to dissolve and wash out polymers generated by the heat of the compression.

Undesired charged gas-adsorber bed interactions sooner or later lead to deterioration of the adsorber bed. Consequently, adsorber bed capacities will be reduced and the pressure drop will increase; both are caused by the formation of polymerisates and coke. Figure 11

Dryer operating pressure, bar 35 15Operating temperature, °C 15 15Water content of charge gas, mg/STDm3 450 840

Charged gas water content as a function of operating pressure

Table 1

47% 50%

3%

>4.5 years2.5-4.5 years<1.5 years

18%

50%

32%

Cycle times too shortPressure-drop issuesVessel inspections

11%

81%

8%

Kept employed dessicantChanged technology for general improvementChange of targeted contaminants

80

100

90

70

60

Bed d

ete

riora

tion d

ue t

o a

gein

g,

%

0.19

0.21

0.17

0.15

0.18

0.20

0.16

Pre

ssure

dro

p,

bar

500 6 12 18 24 30 36 42 48 54 60

Months of operation

Pressure dropBed deterioration

Figure 11 Ethylene plant charged gas dryer bed aging and pressure drop increase

Figure 12a Desiccant life time in ethylene plants

Figure 12b Reasons for changeout Figure 12c Consequences of the changeout

a b

c

It is run at high pressure over an operational plant period of 60 months, equalling about 650 regen-eration cycles. After 650 cycles, the adsorption bed capacity has been reduced to nearly 70% and the pressure drop over the bed has increased from 0.15 bar to 0.21 bar.

According to an AIChE survey,12

operators from gas (60%) and liquid (40%) processing ethylene plants reported the observations shown in Figures 12a to 12c.

The design philosophy for a

gas grace.indd 11 11/03/2014 09:45

18-20 bar for the gas phase, and 16-28 bar for the liquid phase, with temperatures typically in the range 35-50°C. The best desiccant for this drying step, like the second drying step, is a 3A type zeolite with espe-cially low polymerisation and coking characteristics. Such a desic-cant can be used for gas and liquid phase operations. The regeneration gas and the regeneration tempera-ture are the same as for the above charged gas dryer.

The condensate dryerDepending on the technology chosen, hydrocarbon condensates generated at the fourth compres-sion stage have to be separated in the knockout drum upstream of the charge gas dryer. These condensate dryers operate at pressures in the range 30-35 bar, at temperatures of 12-15°C. Typically, two dryers are required: one in service, the other in regeneration or standby mode. The state-of-the-art adsorbent is a special 3A type zeolite. The hydro-carbon condensates contain oligomer and/or polymer precur-sors, which are formed in the upstream compression section. Consequently, the selected adsor-bent should exhibit little tendency to coking and polymerisation. Depending on cycle times, which are 24-48 hours, the lifetime may reach fi ve to six years. The conden-sate is saturated with water, depending on composition and operating conditions. The quality of the upstream separator must ensure that practically no traces of free water reach the dryers. If free water is expected always to be present in the condensate, a layer of macropo-rous silica gel is installed in the bed inlet to guard the zeolitic adsorber bed against free water. The mois-ture concentration in the outlet is specifi ed to be below 1 ppmv. This is to avoid the formation of hydrates in downstream piping and equipment. The regeneration gas and the regeneration temperature are the same as for the dryers.

The hydrogen dryerIn addition to methane, hydrogen off-gas from the ethylene plant’s cold box contains carbon monoxide


any freezing or gas hydrate forma-tion in the ethylene splitter. The secondary dryer runs in a one-vessel mode with a service time of 20 days prior to regeneration. Consequently, during regeneration the secondary dryer step is bypassed. The desiccant of choice for the secondary dryer is a type 3A zeolite with especially low polymerisation and coking charac-teristics. Such desiccant should exhibit low co-adsorption for acety-lene and olefi ns. Depending on the employed technology and the selected separation sequence of the fi rst tower, the operation conditions are: pressures of 20-30 bar and temperatures of -30-30°C. The regeneration gas and the regenera-tion temperature are the same as for the charged gas dryer.

The MAPD converter dryerThe MAPD converter dryer is posi-tioned between the methyl acetylene and propandiene hydro-genation unit (MAPD) and the ethylene splitter to remove any moisture formed during the hydra-tion of the C3 cut. This dryer is needed to avoid any water content in the fi nal product, which is espe-cially important for naphtha based crackers. Typically, two vessels are installed, one in service, the other in regeneration mode, for both liquid and gas phase operations, depend-ing on the chosen hydrogenation technology. The moisture content of the inlet fl uid is typically specifi ed at <100 ppmv. For gas and liquid phases, the dryer outlet fl uid mois-ture contents are specifi ed to be below 1 ppmv. The corresponding operating pressures are in the range

charged gas dryer depends on the contractor:• Main/guard bed confi guration (Lummus, Technip Stone & Webster): the moisture analyser is installed between beds. The guard bed height is typically sized for six to eight hours of additional adsorp-tion time, allowing operators to take action in case breakthrough conditions occur. An additional function of the such guard bed is to adsorb trace contaminants such as ammonia and methanol• The single bed (Linde) adsorbent section below the moisture analyser functions as a guard bed. In this design, one support grid has been omitted and in the same available volume more adsorbent material can be installed. Furthermore, this concept allows an easier loading of the adsorbent material. Both options are shown in Figure 13: • Typical arrangements: two or three vessels in total, one or two in adsorption, one in regeneration mode• Flow: downwards during adsorp-tion, upwards during regeneration• Regeneration gas: methane frac-tion from demethaniser overhead, also utilised for all subsequent dryers. Typical regeneration gas inlet temperature range is 210-230°C• Adsorption time typically 48-64 hours.

Typical lifetimes of low pressure operated units are less than fi ve years, while high pressure units may achieve more than fi ve years of operation.

The secondary dryer The secondary dryer is positioned between the acetylene converter and the ethylene splitter. Its purpose is to remove moisture formed in the acetylene converter during the selective hydrogenation of C2 acetylenes. This is caused by the presence of free oxygen from decomposed oxygenates and/or oxygen ingress into the ethylene unit. A typical specifi ed moisture load for the inlet gas is below 20 ppmv. Licensors offer designs with a limit of 50-100 ppmv of moisture as a safety margin. The targeted moisture concentration of the outlet gas is below 1 ppmv, thus avoiding

x = 0.8 to 1.0 mm

x x

Figure 13 Typical design layouts for charged gas dryers. Left: main bed/guard bed. Right: single bed

gas grace.indd 12 11/03/2014 09:45

www.eptq.com

are zeolitic compounds with very promising adsorption characteris-tics. Unfortunately, these materials are still either difficult to produce, or thermomechanically or chemi-cally not stable enough to serve as adsorbents in ethylene plants.

The shale gas boom leading to the erection of megacrackers is expected to rely on established adsorption technology. Specially developed 3A type zeolites will continue to play a dominant role in this industrial application.

References 1 True W R, Oil & Gas Journal, July 2013, 90-103.2 Falqi F H, The Miracle Petrochemicals, Olefins Industry: An In-Depth Look at Steam Crackers, 7-8, Universal-Publishers, Boca Raton, Florida, USA, 2009.3 Kniel L, et al, Ethylene Keystone to the Petrochemical Industry, 1980, 78-107, Marcel Dekker Inc. New York, NY. 4 Kapur S, Cracking Technology for the Production of Ethylene, Chapter 6.1, Handbook of Petrochemicals Production Processes, McGraw-Hill Professional, 2004. 5 Zimmermann H, et al, Ethylene, Ullmanns Enzylopädie der Technischen Chemie, 2007, 34-47, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.6 Knaebel K S, Adsorbent Selection, Adsorption Research Inc, www.adsorption. com/publications/AdsorbentSel1B.pdf (June 2004), 2003.7 Breck D W, et al, J. Am. Chem. Soc., Vol. 78, No. 23, Dec 1956, 5967. 8 http://helios.princeton.edu/zeomics/ cgi-bin/view_structure.pl?src=iza&id=LTA& token=e8279dfc54f813ef8f88b9baad8 bcc729 Rode E J, et al, Molecular sieve adsorbent properties and their impact on cracked gas drying, AIChE Spring National Meeting 1997, Houston, TX. 10De Haan S, et al, Overview of Cracked Gas Compressor Fouling: Theories and Practices, Presented at: 2001 Spring National Meeting, 23-26 April 2001, Houston, TX.11Snider S, Olefin Plant cracker Gas Compressor Fouling, PTQ Gas, 2007, 36-40.12 O’Brien B, et al, Cracked Gas Drier Survey, Spring National Meeting, 21-25 March 2010, San Antonio, TX.

Hans Hoefer, a materials scientist by education, worked for almost 27 years for WR Grace in the Adsorbents Business as Manager for Research & Development, Technical Customer Service and Market Development.Vassilios Zafirakis is Manager of Technical Customer Services Adsorbents with WR Grace, Worms Germany. He holds a Master’s in process engineering from the University of Stuttgart, Germany.

(CO). Since this acts as an inhibitor for palladium and silver hydro-genation catalysts, untreated hydrogen off-gas cannot be used for hydrogenation (acetylene converter, MAPD converter). For this reason, the CO has to be removed. There are basically two established options (refer to process flow diagrams):1. A methanation reactor for low grade hydrogen2. PSA for high grade hydrogen purification.

In the first case, CO in the hydro-gen rich stream is catalytically converted to methane:

CO + 3H2 -> CH

4 + H

2O

H2O is a poison for the hydro-genation catalyst and has to be removed. A TSA dryer equipped with a 3A or a 4A type zeolite has a typical cycle time of 24 hours. The operating pressure is 6-30 bar at 15-30°C. The regeneration tempera-ture is 240°C. The regeneration gas is the demethaniser overhead. If the feed is free of olefinic and acety-lenic compounds, the lifetime may be between five and 10 years.

In the second case, the hydrogen rich stream is purified by a three-layer PSA unit consisting of: 5A type zeolite to remove nitrogen and CO; activated carbon to remove methane and CO2; and activated alumina to eliminate H2O and NH3. The adsorption pressure is 15-30 bar. The desorption pressure is typically lowered by factors of 10-13 but is always kept slightly above atmospheric pressure. The operating temperature is 15-35°C. The high grade hydrogen has a typical purity of 99.5 vol% with concentrations of CO <10 ppmv, CH4 <10 ppmv, and N2 <10 ppmv.

OutlookThe adsorption technology discussed here has its roots in the early 1950s. Over the years, a lot of progress has been made in the adsorbents’ performance and economical feasibility. Although nowadays there are about 160 differ-ent zeolite structures, none of those could replace the type 3 A zeolite for ethylene drying. Certainly there

Our knockout drum needs a mist eliminator!

Can we add one without overhauling the vessel?

IT’S A MORE COMMON problem than you might think. A vertical knockout drum removes free liquid from a certain gas stream. But at the time the plant was built, a mist eliminator was not considered necessary. Now mist is carrying over and causing trouble downstream. There is no manway, so adding a conventional mist eliminator would require cutting the vessel open.

Instead, if the fl anged exit nozzle is at least six inches in diameter, you can simply hang a patented MistFix® mist eliminator inside, securing it between the fl anges. Even when there is a manway, MistFix avoids entry, welding, and recertifi cation in retrofi ts. For new and old vessels, it drastically cuts installation cost and downtime.

Phone: 713-434-0934 • Fax: [email protected]

24-hr Emergency Service: 281-716-1179

MistFix® insertion mist eliminator

Visit our new website atwww.amacs.com

gas grace.indd 13 11/03/2014 09:45

http://www.digitalrefining.com/56,sponsors,AMACS_Process_Tower_Internals.html

Organised by CRU Events

3-6 November 2014, Marriott Rive Gauche, Paris, France

Simultaneous English/Russian interpretation/Синхронный перевод на английский/русский

See you in Paris to celebrate our 30th Anniversary!

Register today at www.sulphurconference.com

Organised by CRU Events

Operator/Producer

50%discount

Early bird rate just

€1405

Oil | Gas | Fertilizers | Metallurgy | Industrial uses

Sulphur 2014The world’s leading industry event

2014 Call for papers

Sponsored by:

Official publications:

Case studies demonstrating operational experience in:- Process and efficiency improvements- Advances in environmental efficiency and strategies

for compliance- Equipment improvements and optimisations- Operational problems and solutions- Best practice health and safety initiatives- Application of improved technology or catalysts- Restoration, revamping or maintenance

Technology innovations and improvements:- New technology developments, which, if not proved

so far, offer the realistic prospect of economic advantage in plant operations

- Technology innovations that improve environmental or health and safety regulation compliance

Interested in presenting a paper in Paris? Submit your abstract by 14 April 2014 online at www.sulphurabstracts.com or visit www.sulphurconference.com for further details

Sulphur is an excellent place to see the latest developments in both sulphur & sulphuric acid with great presentations as well as a great place to meet both new and existing equipment suppliers.Research Engineer, Haldor Topsøe A/S

“ ”

48_sulphur.indd 1 06/03/2014 13:49

Virtual commissioning of a gas handling system

Gas cargo such as liquid natu-ral gas, ethane, ethylene and propylene is transported in

liquefied form under cooling temperatures and pressures. Typically, pressures are low, which leads to very low temperatures. Although the cargo tanks are well insulated, gas continuously evapo-rates from the liquid cargo. This gas is called boil-off gas and has to be reliquefied during maritime transport in the reliquefaction system.

The reliquefaction system by TGE Marine Gas Engineering for five new vessels of Navigator Gas ship-ping company was designed with UniSim Design. First, a stationary simulation model was used for the basic design. Later, a dynamic simulation model was developed for the design of the automation and control system.

There are many differences between automation systems that are installed on a vessel and auto-mation systems of a production plant. For instance, it is much more difficult to provide support and maintenance on a vessel than on land. As a conse-quence, sea based automation systems have to be designed, verified and tested very care-fully. The requirements in this field are to some extent similar to the requirements in the automobile and aviation industries.

In order to develop auto-mation systems that are safe, reliable and error free, auto-mobile and aviation industries rely on hardware in the loop simulation. With

Gas reliquefaction units were designed by means of stationary and dynamic process simulation models

RAINER SCHEURING Cologne University of Applied Sciences HANS-CHRISTIAN HAARMANN-KÜHN TGE and JÜRGEN ESSLER TGE Marine Gas EngineeringTORSTEN FELIX Flemming Automationstechnik MICHAEL BRODKORB Honeywell Process Solutions

this technology, the automation system can be verified within a simulation based test environment.

TGE has decided to use hardware in the loop simulation in the design of the automation system of the new cargo handling system, too. This allows not only the develop-ment of an error-free automation system but also a substantial reduc-tion in commissioning time.

Reliquefaction systemAbout 50% of the existing fleet of liquefied ethylene (LEG) carriers uses TGE designed cascade relique-faction units to cool down cargo and maintain tank pressure. Liquefied ethylene is usually trans-ported using semi-pressurised, fully refrigerated vessels that are equipped with IMO type C tanks (see Figure 1) and cascade refrigera-tion systems.

The basic technology of the cascade refrigerant systems have remained unchanged for years. The cargo tanks are cooled by an open cooling system compressing the

BOG in a two-stage process and condensing it against a refrigerant (see Figure 2). Two crank, oil-free piston compressors usually serve as cargo compressors. The refrigerant (propylene or R404A in place of R22) is compressed by oil injected screw compressors and condensed against seawater.

To date, the biggest ethylene carriers in operation are 22 000 m³ vessels built by Jiangnan Shipyard for Navigator Gas. A new genera-tion of these vessels with 21 000 m³ capacity is under construction and the first of five vessels are to be delivered in 2014.

The vessels are designed to carry 21 different LPG, chemical gas and chemical cargoes. For 15 of them (including ethylene, propylene, ammonia, VCM and butadiene), refrigeration is provided. The vessels are capable of transporting up to three different grades, two of which may be cooled at the same time. Operating conditions and modes vary widely depending on the specific cargo.

A challenge in the design of the reliquefaction units was to improve perfor-mance and operability in order to meet increased requirements for loading rates and cooling down times. In addition, the same reliability and robustness of the units, which have operated successfully for 14 years, had to be maintained. As an example, cooling down capacity for ethylene in one-grade operation is 50% higher for the


Figure 1 5500 m³ bi-lobe tank under construction at Jiangnan heavy industry site in Shanghai

gas honeywell.indd 1 11/03/2014 11:51


Aspen Hysys Dynamics are based on so-called first principle process modelling engines that allow realis-tic modelling of the transient behaviour of processes typically found in the oil and gas and chemi-cal industries. In order to create a process model, the user selects readily available components and thermodynamic packages to define physical properties and phase equilibria of the system and then creates a flowsheet by adding and linking generic unit operation models (such as pipes, vessels, pumps and distillation columns) and control equipment (valves, PIDs, and so on). The resulting model can be initialised to a specific initial condition and run through different predefined scenarios as part of a dynamic simulation study.

Dynamic simulation studies are a standard tool in the process indus-tries for analysing and optimising transient process behaviour. Application examples for operabil-ity or safety studies include dynamic flare load estimation in

new systems. In order to meet all requirements, compressors of higher capacity and new modes of operation have been introduced.

Dynamic plant simulation and design of automation systemDynamic Process Simulators like Honeywell UniSim Design, Invensys Dynsym, or AspenTech

Cargo liquid separator LPG

condensor

Refrigerant liquid separator

Refrigerant compressor

Refrigerant condensor

Refrigerant collecting vessel

Refrigerant economiser

Cargo vapour compressor

Cargo condensor

Cargo receiver

Cargo tank

Cargo vapour

Out

Sea water

In

Out

Sea water

In

Figure 2 Reliquefaction system

C

M

Y

CM

MY

CY

CMY

K

OGT_124_ProTreat_Half_Page_Horizontal_PTQ_Gas_Supplement_2014.pdf 1 3/4/14 8:29 AM


http://www.digitalrefining.com/136,sponsors,Optimized_Gas_Treating.html


Automation system(Siemens S7)

Simple signal simulation(Simit)

Comprehensive plant simulation(UniSim)

Figure 4 HIL simulation of TGE’s gas carrier cargo handling system

Cargocondensor

Refrigerant compressor

Refrigerant condensor

Refrigerant economiser

Refrigerant collecting vessel

Figure 3 Dynamic simulation model of reliquefaction system

refineries1 or onshore gas fields,2 and compressor studies.3

In order to design a control concept for the reliquefaction system, a dynamic simulation model was developed (see Figure 3).

This model served as the basis for the design of a new control concept that makes it possible to operate the reliquefaction system with all cargoes and all operation modes within a single control structure. It is TGE’s first fully automated cascade refrigerant cycle for LEG carriers. In addition, the reliquefac-tion system is kept in a stable and stationary operation point inde-pendent of any load on the system. In order to ensure that the relique-faction system operates in a satisfactory manner in all operating conditions, special attention was paid not only to efficiency but also to robustness of the solution.

Hardware-in-the-loop simulation Hardware-in-the-loop (HIL) simu-lation is a technique that is used in the product development cycle in which real components interact with simulated components.4 Early

applications date back to the 1970s.5 Today, HIL simulation has become an integral component in the devel-opment process of electronic control units especially in the automotive and aviation industries.4,6 In the basic structure of a HIL simulator, the control unit is a real component and the plant is simulated.

Due to the fact that a gas carrier

cargo handling system is installed on a seagoing vessel, activities such as troubleshooting and maintenance are difficult to handle. As a result, the automation system of a gas carrier cargo handling system has

to be designed, verified and tested very carefully. The aim is to have an error-free system. Figure 4 presents the HIL simulation struc-ture of TGE’s gas carrier cargo handling system.

The automation system of TGE’s cargo handling system is based on a Siemens S7 programmable logic controller (PLC).

In order to test the automation system, two simulators were used: Simit and UniSim. Simit can be smoothly integrated into Siemens automation hardware and software infrastructures and supplements

Dynamic simulation studies are a standard tool in the process industries for analysing and optimising transient process behaviour



behaviour of this system was exam-ined for a representative subset of cargos and all load scenarios.

In a final test session, human machine interface, start-up behav-iour, normal operation and several fault scenarios were verified with experienced engineers from the shipping company.

As a result of these inquiries, several logic faults and control algorithm specific implementation errors have been eliminated in front of the tests at the shipyard and the trial run.

ConclusionTGE Marine Gas Engineering has developed a new, more flexible, and more powerful gas carrier cargo handling system. This system is able to handle a high number of different cargos such as LPG, gases and chemicals. For the majority of these cargoes, refrigeration is provided. This function is performed by reliquefaction units that were designed by means of stationary and dynamic UniSim Design process simulation models.

Elimination of errors and debug-ging of automation software is one of the main tasks during the commissioning phase of a cargo handling system. Due to the fact that commissioning time is very limited, the complex logics of the automation system and the dynamic behaviour of the relique-faction units cannot be verified in detail during this phase. Therefore, TGE decided to carry out a detailed HIL simulation study of the gas carrier cargo handling system at an earlier design stage.

In the HIL simulation study the automation system was extensively tested and evaluated. Simit was used for the verification of the complex logics. The S7 code components, which control the central reliquefaction system, were tested by connecting them to a dynamic UniSim Design process simulation model and carrying out a virtual commissioning. As a result, several logic faults and control algorithm specific imple-mentation errors have been eliminated.

in the shipyard. Subsequently, a trial run with one cargo and dura-tion of about one week takes place. It is neither possible to verify the complex logics of the automation system in detail nor to analyse and optimise the dynamic behaviour of the reliquefaction system for different cargos and load scenar-ios. As a consequence, TGE decided to carry out a detailed HIL simulation study of the gas

carrier cargo handling system at an earlier design stage.

Simit provides a test environment for the S7 based automation system. A strong point is its ability to test the logic controls of the S7 PLC. Therefore Simit was used to test and verify the logic controls of the cargo system.

UniSim was employed for a detailed analysis of the S7 code components that control the central reliquefaction system. The dynamic

the automation interface hardware (Profibus DP, Profinet IO). Communication between the auto-mation system and Simit is carried out as in the real system at the field bus level.7

Simit does not provide a first principles modelling engine. Consequently, a dynamic UniSim model was used for testing the reli-quefaction system. In these tests, Simit established communications between UniSim and the S7 PLC. The interface between UniSim and Simit was developed using UniSim Design OLE Automation client-server technology.

Virtual commissioning of automation systemElimination of errors and debug-ging of automation software is one of the main tasks during a system’s commissioning phase. In the case of machine tools, a study from 1997 provides the data for Figure 5.8 The duration of commissioning takes approximately 15-25% of the entire project. About 90% of this period is required for commission-ing the automation system, of which up to 70% are needed for the elimination of software errors (see Figure 5).

Commissioning engineers for gas carrier cargo handling systems still experience the same difficul-ties. In addition, commissioning of these systems has to deal with a core problem: commissioning time is very limited. Some preliminary testing without cargo can be done

Entire project duration

Duration of commissioning

Duration for automation

system

15-25%

Up to 90% for automation

system

Up to 70% for software errors

Figure 5 Fraction of commissioning and project duration for software error elimination

Elimination of errors and debugging of automation software is one of the main tasks during the commissioning phase of a cargo handling system



from Stuttgart University, Germany.Email: [email protected]

Hans-Christian Haarmann-Kühn is the Head of the Engineering Department at TGE Marine Gas Engineering GmbH, Bonn, Germany. He has a degree in process engineering and a PhD from Technical University RWTH Aachen, Germany. Email: [email protected]

Jürgen Essler is a Project Engineer at TGE Marine Gas Engineering GmbH in Bonn, Germany. He has a degree in mechanical engineering and a PhD from Technische Universität Dresden. Email: [email protected]

Torsten Felix is certified PLC engineer and Project Manager at Flemming Automationstechnik GmbH.Email: [email protected]

Michael Brodkorb is Advanced Solutions Consultant at Honeywell Process Solutions in Tarragona, Spain. He has a degree in chemical engineering from University of Dortmund, Germany, and a PhD from Bradford University, UK. Email: [email protected]

4 Wältermann P, Hardware-in-the-Loop: the technology for testing electronic controls in automotive engineering. 6. Paderborner Workshop: Entwurf mechatronischer Systeme, Heinz Nixdorf Institut, Paderborn, 2.-3.4.2009,. www.dspace.de/shared/data/pdf/paper/HIL_Overview_Waeltermann_03_E.pdf.5 Waite W F, Simulation – the past as prologue …probably, ITEA Journal, Mar/Apr 2002, 7-10.6 Ledin J A, Hardware-in-the-Loop Simulation. Embedded Systems Programming, Feb 1999, 42–60. 7 Siemens: SIMIT V7 - open simulation platform for virtual commissioning, www.industry.siemens.com/verticals/global/en/chemical-industries/chemical-industries-products-systems/Documents/SIMIT_Brochure_EN.PDF.8 VDW-Bericht: Abteilungsübergreifende Projektierung komplexer Maschinen und Anlagen. Verein Deutscher Werkzeugmaschinenhersteller, Aachen, 1997.

Rainer Scheuring is Professor of Automation Technology and Control Theory at Cologne University of Applied Sciences. He has a degree in technical cybernetics and a PhD

With the HIL simulation, virtual commissioning of the automation system has been realised. This will significantly contribute to a safe, reliable and error-free gas carrier cargo handling system.

UniSim is a registered trademark of Honeywell International Inc. DYNSYM is a registered trademark of Invensys Inc. Aspen Hysys Dynamics is a registered trademark of Aspen Technology Inc. SIMIT is a registered trademark of Siemens AG

References1 Gruber D, Leipnitz D-U, Sethuraman P, Alos M A, Nogues J M, Brodkorb M, Are there alternatives to an expensive overhaul of a bottlenecked flare system?, PTQ, 93-95, Q1 2010.2 Panigrahy P, Balmer J, Alos M A, Brodkorb M, Marshall B, Dynamics break the bottleneck, Hydrocarbon Engineering, 93–96, Sept 2011.2 Nugues J M, Brodkorb M, Feliu J A, How can dynamic process simulation be used for centrifugal compressor systems, Hydrocarbon Engineering, 92–98, Aug 2012.

4th OPPORTUNITY CRUDES CONFERENCE

Downstream Meets Upstream: Right Crudes to Make Products in Demand

WESTIN OAKS HOTEL | HOUSTON, TX (US)| SEP. 14-16, 2014

This international conference addresses the difficulties and benefits of processing aromatic opportunity crudes and paraffinic light oil alone and together as well as acidic or high TAN crudes.

REGISTER TODAY at opportunitycrudes.com/houston2014 to take advantage of the early bird discount.

We have assembled a diverse team of technology experts and business strategists to share their experiences and insights into many urgent issues facing refiners worldwide:

Accelerated US shale/tight oil output, influx of NG condensate, expanding Canadian oilsand output, and increased processing of heavy oil outside the US

Immiscibility concerns when blending aromatic heavy crudes and paraffinic tight oils into so-called "dumbbell" crudes

Corrosion, fouling, and asphaltene deposition in railcars, pipelines, terminals and refinery units

Uncertainty over FCC and coker operations due to a shift in crude types, a decline in gasoline demand, and a naphtha “surplus” from tight oil processing

Seeking advanced residue and tight oil processing technologies to produce more middle distillates and petrochemicals amid shrinking fuel oil demand

Our speakers come from over 25 companies, including Albemarle, Baker Hughes, BASF, Cameron, Canadian Energy Research Institute, Center for Strategic and International Studies (CSIS), Ecopetrol, EnSys Energy, Haverly Systems, Intertek, Invensys Spiral, Johnson Matthey-INTERCAT, KBC Advanced Technologies, Nalco Champion, Schneider Electric, Soluble Solutions, Turner Mason, and others.

Don't miss out on opportunities to exchange ideas, share experiences, and network with professionals and executives working in upstream, midstream, and downstream of the opportunity crudes and tight oil business.

Host HYDROCARBON PUBLISHING COMPANY Translating Knowledge into ProfitabilitySM Sponsors Media Promoztional Partners


Protecting compressors with dynamic simulation

Integrated dynamic simulation of gas processing and petroleum refining processes is vital in the

prevention of catastrophic equip-ment occurrences. The protection of compressors from mechanical fail-ure is essential to maximise operating time and ensure safer operations.

This article highlights how dynamic engineering software successfully meets tighter product specifications through improved understanding of plant operability issues and ensures faster, safer plant start-ups to help avoid unplanned downtime. The results deliver maximum plant availability and productivity.

Better understanding leads tobetter plant controlUnderstanding the process is the first step in the design of an effec-tive control strategy. To prescribe a solution, the process control engi-neer must analyse all variables, process dynamics and unit interac-tions. First principle models (steady state and dynamics) provide a better understanding of the process dynamics and its interactions, enabling engineers to evaluate and tune strategies before implementa-tion. Using dynamic simulation early in the design phase can help identify important operability and control issues and influence the design accordingly.

Compressors used to increase the pressure of a gas are necessary in many process designs. They are highly specific pieces of equipment and often custom designed and, therefore, expensive to purchase. Compressors operate with fast

Dynamic simulation supports reliable operation of compressor installations to deliver maximum plant availability and productivity

NICHOLAS BROWNRIGG AspenTech

dynamics, meaning that a small disturbance can disrupt a compres-sor away from steady state performance very quickly.

When steady state performance is interrupted, a compressor can expe-rience a phenomenon known as surge. This can lead to malfunction-ing or, worse, induce major compressor damage requiring equipment replacement. In addition, a compressor surge can harm indi-viduals working near compressors or release noxious gases into the surrounding atmosphere.

Using dynamic simulation can help to predict compressors’ devia-tion in behaviour from steady state that could cause compressor surge. Dynamic control schemes can also be modelled to limit steady state deviations if a disturbance is expe-rienced in the plant.

The first step in any dynamic simulation is obtaining a steady state model. A steady state compressor model shows undis-turbed compressor operation or, in other words, how the compressor would ideally operate if uninter-rupted, given the process parameters. To generate the steady state compressor model in process simulation, the inlet streams of the compressor model must be defined, as well as one of three compressor specifications: the change in pres-sure across the compressor, the compressor’s pressure ratio or the energy supplied to the compressor.

Model based operations improve decision support and safetyCompressors are vital to gas processing and similar operations, so it is important that models

include a built-in response to potential disturbances, such as a compressor surge. Dynamic compressor models can optimise design and operational perfor-mance to achieve the following:• Analyse a range of process disturbance scenarios• Add a proper control scheme to appropriately respond to process disturbances• Perform process safety studies• Finely tune process parameters to achieve improved production results.

Process engineers use AspenTech’s aspenONE solutions to achieve model based operations for decision support and safety. In particular, Aspen Hysys Dynamics extends Aspen Hysys steady state models into dynamic process models, enabling design and verifi-cation of process control schemes, safety studies, relief valve sizing and rating, failure analysis and development of start-up, shutdown and operating mode changes. Engineers can create steady state and dynamic models all within the same simulation environment, which saves significant time and effort (see Figure 1).

Extensive numerical and graphical results are generated for dynamic models in Aspen Hysys Dynamics, which allows for in-depth analysis to be performed on each completed simulation (see Figure 2).

Many compressor disturbances can be explored, including decreas-ing or increasing the feed rate to the compressor and modifying the composition or physical properties of the compressor feed stream, among others. Control schemes and parameters can also be modified,


aspentech.indd 1 12/03/2014 12:49


behaviour, many production processes in the oil and gas, energy and chemicals industries will expe-rience signifi cant improvements in control and safety. Knowing the dynamic response of a compressor is important when designing or operating a process, and engineer-ing dynamic software is a tool that allows users to perform dynamic simulations to obtain this informa-tion. Engineering modelling software provides multiple outlets to view and tabulate the dynamic response for equipment and controllers, including strip charts, tables and compressor curves. When Hysys and Hysys Dynamics are properly utilised conjointly, compressor surge can be avoided. Therefore, through better under-standing of plant behaviour, companies will save equipment costs.

Verifying safe operation of a plant revampPetronas, a Malaysian oil and gas company, is one of the world’s largest and most profi table compa-nies. Petronas operates a liquefi ed natural gas (LNG) plant that under-went a revamp. While designing the revamp, Petronas turned to Aspen Hysys Dynamics to verify that the new operating conditions and control schemes, as well as

yielding more or less robust compressor control, depending on the user’s requirement. It is a powerful tool that can be utilised to ensure safe and proper functioning of a process’s compressors.

Engineers can broaden their abil-ity to design safer, more operable plants, without over-designing, by learning to do the following: • Use data visualisation to help optimise performance • Implement and test control schemes• Schedule events to study safety scenarios, start-up and more• Switch steady state models to dynamic mode for greater accuracy and fl exibility.

Reported benefi ts using dynamic modelling can generate $15 million savings through improved and faster start-up procedures. Also, the avoidance of over-designing relief systems can achieve $10 million in capital cost savings, improved safety through better operational procedures, better control system design and proper relief-valve sizing delivers enormous benefi ts. Operators can achieve better design decisions through detailed analysis of the trade-offs between process operability and process integration.

Safety and controllability mitigates equipment malfunction By understanding dynamic plant

9

3

2

6

5

N2

Q-23KA00

723L1005-03

4

18

Q-100

VLV-102

V-100

23FE1015

PIC-100

dP scrubber

dP and vol 23Lxxxx

Bath tube strainer

23KA001

23KA001: surge controller

E-100VLV-101

VLV-100

TEE-101MIX-100

Molar flow kgmole/hkPAkPA%

19001.073192.7

100.00

Pressure dropFeed pressurePercentage open

23FE1015

* Actual feed volume flow

Compressor speed rpmkWACT_m3/hkPA

62853725

26159.5183.6

PowerCapacity *Feed pressure

kPA1220Product pressure

23KA001

Figure 1 Steady state model in Aspen Hysys

Figure 2 Simulation of a compressor installation in Aspen Hysys Dynamics

TRF-1

aspentech.indd 2 12/03/2014 12:49


start-up and shutdown procedures, would not lead to compressor surge in one of the plant’s many compressor units.1

In Hysys Dynamics, models were constructed that represented actual equipment operation, piping arrangements and potential control schemes. Of note, a realistic emula-tion of a controller used by Petronas for anti-surge control in compressors was replicated. In addition, vendor predicted and as tested operating curves for the axial and centrifugal compressors in the plant were input into the package for enhanced prediction of compressor behaviour.

During revamp studies, a variety of new operating conditions and failure scenarios for the plant was considered. The use of dynamic simulation allowed for the testing of the operating conditions to ensure that they would not cause disruptions with equipment perfor-mance and, in particular, compressor surge. Failure scenario analysis enabled Petronas to view the operating limits of the compres-sors along with specific circumstances that would directly lead to surge in order to better avoid them. Failure scenario analy-sis also helped to confirm that relief valves were adequately sized to handle peak relief loads.

While performing dynamic simu-lation, it was elsewhere discovered that a propane compressor did, in fact, experience surge upon the closure of suction valves on a chiller and also during a trip of a cryogenic heat exchanger. This risk was quickly mitigated since it was uncovered early in the design process, before implementation of equipment. In all, over 45 different scenarios, along with the start-up and shutdown conditions of the plant, were simulated and analysed in an iterative process utilising the event scheduler feature inside Hysys Dynamics. Using the pack-age to carry out dynamic simulation for the revamp of the plant ultimately lent a better insight into the dynamic performances that new plant designs would exhibit.

By knowing where potential dangers in the process lay, Petronas

was able to better size and operate equipment in order to avoid the occurrence of a catastrophic event.

Operability analysis for emergency system responsesWood Group Production Services Network (Wood Group PSN) is a consulting group that assists oil and gas companies in attaining better performance and production, from process development to oper-ational diagnostics. Due to the complex nature of safety systems, performing safety system analysis is a task that Wood Group PSN often executes.

Safety systems are often designed more conservatively than required in order to protect against worst case scenarios. However, the worst case scenario calculation is a poor representation of actual process circumstances and it does not offer any insights regarding how a process behaves on its way to reaching the worst case. Safety systems’ performance and sizing can be vastly improved through dynamic simulation to ascertain answers into how processes actu-ally respond to disturbance. For this reason, Wood Group PSN was tasked with creating dynamic simu-lations to model the response to steady state disturbance of emer-gency systems present in a plant with the goal of ultimately enhanc-ing the safety of the process. To accomplish this, the company chose to use Hysys Dynamics.

A comprehensive model of the plant’s safety system was devel-oped to investigate equipment failures as conditions moved away from the steady state and to evalu-ate alternative process configurations towards bettering safety system effectiveness. Compressors were heavily featured in many of the dynamic models created, and the performance and efficiency curves for the compres-sors, as well as the inertia and friction profiles for each compres-sor, were included to better tailor the model to fit the actual process.

Studying failure scenarios allowed Wood Group PSN to view which circumstances initiated compressor surge and to what

degree. Accordingly, the safety margins of the compressor units were increased until confirmed to protect against the occurrence of surge. Using dynamic simulation enabled the safety margins to be increased up to the critical level of protection based on predicted process responses and eliminated the need for over-design to provide a blanket protection.

A colour scheme was imple-mented on the dynamic simulations to easily identify when material became trapped and caused pres-sure build-up when malfunctions occurred. Identifying where block-ages in operation take place helped to prevent surge, since flow disrup-tions is one of the main causes for breakdown. Additionally, using dynamic simulation eliminated the need for the installation of a hot gas bypass that served to prevent compressor malfunctioning. This reduction in design saved over $3.2 million in capital expenditure for the plant. The requirement of the hot gas bypass is an example of the over-design that can materialise when worst case scenario analysis is used for the creation of safety systems and how the use of dynamic simulation can help determine where unnecessary components of the safety systems exist.

The models created in Hysys Dynamics were also able to serve as a basis for operator training and on-going development in the plant. With an improved knowledge of how the plant runs, operators were better at steering the process to avoid perturbations from steady state and how to best remedy situa-tions that have left steady state operation.

References1 AspenTech Global Conference, 2010.2 AspenTech User Conference, 2009.

Nicholas Brownrigg is Dynamics Product Manager, AspenTech.

aspentech.indd 3 12/03/2014 12:49

2014 Speakers include:

Boost your refi ning profi tability through operational excellence and increased product yields

8th Annual

Global Refi ning Summit31 March-2 April 2014

Hesperia Tower, Barcelona, Spain

Save £300* when you book your place by 31 March 2014

refi ningsummit.com/PTQ

*The offer is for end-users only and is not available to equipment manufacturers and solution providers.

[email protected] | +44 (0) 20 7202 7690 | refi ningsummit.com

Lars Thorstholm Manager StrategyStatoil ASA

Giacomo Rispoli Executive Vice Presidenteni S.p.A.

Rustem GimaletdinovHead of Department, Technical & Engineering Excellence, Global Refi ning, Petrochemicals & Gas Processing JSC LUKOIL

Rob Overtoom Technology Manager Pearl GTL, QatarShell

Marcus Lippold Vice President, Business Strategy and Regulatory and Public AffairsMOL Group

Mohamed Saad Al-KubaisiTechnical ConsultantQatar Petroleum Refi nery

global summit.indd 1 04/03/2014 11:08

Your objectivesin focus

Make the most of today’s and tomorrow’schallenges with leading-edge solutions from Axens

- Clean and alternative fuel technologies - Petrochemicals - Energy efficiency- High performance catalysts & adsorbents - Revamps

Single source technology and service providerISO 9001 – ISO 14001 – OHSAS 18001

www.axens.net

210/297.qxd:A4 13/08/10 12:12 Page 1

www.ptqenquiry.com for further information

axens.indd 1 13/9/10 13:27:18

http://www.digitalrefining.com/57,sponsors,Axens.html

2014 gas__

Documents

Transcript of 2014 gas__