IS YOUR FACILITY PERFORMING?

A look at the Richard A. Reynolds Groundwater Desalination Facility

Presented by: Gabriela Handley

SUMMARY

Facility Background

Data Normalization

Historical Data

Cleanings

Current Operation

Long-term Shutdown and Membrane Preservation

BACKGROUND: R.A.R. FACILITY

Maximum daily production of 4 mgd sourced from six brackish groundwater wells.

Final product water quality goals: TDS<500 mg/L, chloride <250 mg/L, and manganese <0.05 mg/L.

Three RO trains in a 20:10 array at 81% recovery.

Bypass blend water is treated through an iron and manganese removal system.

Current membranes are Toray TMG20-400C installed January 2010.

Expansion will add five new brackish groundwater wells and increase capacity to 10 mgd.

DATA NORMALIZATION

Calculation of values that describe membrane performance at design conditions even when system is not operated at design conditions

Removes influence of variable operating conditions Temperature

Flow

Recovery

KEY NORMALIZED PARAMETERS

Specific Flux (Normalized Permeate Flow)

Normalized Differential Pressure

Normalized Permeate Conductivity

NORMALIZED DATA REVIEW

High normalized permeate conductivity Poor water quality

Indicates damaged or deteriorated membrane

Low specific flux Low permeate flow

High feed pressure

Indicates problem on membrane surface

High normalized dP Increased energy costs

Indicates problem in feed/concentrate channel

SCALING AND FOULING OF ROMEMBRANES

Fouling is a more generic term: Can include suspended solids, colloids, microorganisms, and organic

chemicals. More difficult to predict Can occur at front end (lead elements) or back end or overall

Scaling refers to formation of a chemical “scale” similar to what occurs in a water heater, boiler, or pipe. Typically scales are calcium carbonate, sulfate scales of calcium,

barium, or strontium and silica Scaling is more or less predictable based on water chemistry Generally occurs at back end of RO system (tail elements).

SYMPTOMS AND CAUSES OF FOULING Symptoms:

Permeability loss (decrease in specific flux)

dP increase – especially in the first stage

Causes: Fine particulate- Iron and manganese very common

Operation at elevated flux or recovery

Changing feedwater quality

SYMPTOMS AND CAUSES OF SCALING Symptoms:

Decreased tail end permeability Increased tail end salt passage dP increase or decrease in tail end (depends on type of precipitant)

Causes: Ineffective antiscalant Loss of antiscalant or inadequate dose Insufficient acid addition Operation at elevated recovery

SPECIFI

C

FLUX

NORMALIZED

DP

STAGE

NORMALIZED

DP

NORMALIZEDPERMEATECODUCTIVITY

PAST CLEANINGSTrain 1 Train 2 Train 3

May 2011 1st stage only AvistaP303

1st stage only AvistaP303

1st stage only AvistaP303

May 2012 1st stage only AvistaP303

1st stage only AvistaP303

1st stage only AvistaP303

May 2013 1st and 2nd stage Avista P303

1st and 2nd stage Avista P130

1st and 2nd stage Avista P130

May 2014 1st stage twice 2nd

stage once AvistaP130

1st stage twice 2nd

stage once AvistaP130

1st stage twice 2nd

stage once AvistaP130

December 2014 1st stage twice 2nd

stage once AvistaP130

1st stage twice 2nd

stage once AvistaP130

1st stage twice 2nd

stage once AvistaP130

June/July 2015 1st stage twice 2nd

stage once AvistaP130

1st stage twice 2nd

stage once AvistaP130

1st stage twice 2nd

stage once AvistaP130

CLEANING PROCEDURE: STAGE 1,BLOCK A

Fresh batch of chemicals

Circulate in Block A for 2 hours

Soak in Block A for 30 minutes

Recirculate in Block A for 30

minutes

Neutralize/dispose of chemicals

Rinse the membranes three

times

Prepare fresh batch of

chemicalsCirculate in Block

A for 2 hours

Soak in Block A for 30 minutes

Recirculate an additional 30

minutesKeep chemical for cleaning of Block B

CLEANING PROCEDURE: STAGE 1,BLOCK B

Circulate cleaning chemical left from

2nd cleaning of Block A for 2 hours

Soak in Block B for 30 minutes

Recirculate in Block B for 30

minutesNeutralize/dispose

of chemicals

Neutralize/dispose of chemicals

Rinse the membranes three

times

Prepare fresh batch of cleaning

chemicalsCirculate in Block

B for 2 hours

Soak in Block B for 30 minutes

Recirculate in Block B for 30

minutes

Keep chemicals for cleaning of

stage 2

CLEANING PROCEDURE: STAGE 2

Circulate cleaning chemical left from cleaning of Block

B

Soak in stage 2 for 30 minutes

Recirculate in stage 2 for 30

minutes

Neutralize/dispose of chemicals

Rinse membranes three times

Full train cleaning complete

CLEANING RESULTS

Initial cleaning interval was 1 year and cleanings were performed each May with Avista P303 cleaner.

The third cleaning of train 1 in 2013 yielded less effective results which prompted a cleaning investigation and element autopsies

Cleaning trials were performed by Avista and a new cleaner for the third cleaning (2013) of trains 2 and 3 was selected, P130.

SPECIFI

C

FLUX

NORMALIZED

DP

STAGE

NORMALIZED

DP

NORMALIZEDPERMEATECODUCTIVITY

FIRST STAGE FOULING EVIDENCE

FIRST STAGE LEAD ELEMENT AUTOPSY 2013

SECOND STAGE TAIL ELEMENT AUTOPSY 2013

CLEANING RESULTS

Increase in fouling rate in late 2014 prompted a second cleaning in December 2014.

Following the December cleaning, the data was normalized and possible causes for the increase in fouling rate were investigated.

The sixth cleaning was scheduled for June/July 2015 and elements from the first and second stage of train 3 were removed prior to the cleaning for wet testing.

Following the cleaning of train 3, the same elements were again removed from the system and wet tested.

SPECIFI

C

FLUX

NORMALIZED

DP

STAGE

NORMALIZED

DP

NORMALIZEDPERMEATECODUCTIVITY

WET TEST RESULTS: TRAIN 36/19/15 Avista Wet Test Pre-Clean 7/1/15 Avista Wet Test Post-Clean

First Stage Vessel #14

First Stage Vessel #14

Position Delta psiNormalized

FlowNormalized

Reject %Weight

lbs Comments Position Delta psiNormalized

FlowNormalized

Reject %Weight

lbs Comments1 25 3.06 98.3 43 Iron/Seperated Vexar 1 6 5.73 98.4 Iron/Seperated Vexar2 13 3.61 95.7 40 Iron/Seperated Vexar 2 5 5.67 97.0 Iron/Seperated Vexar3 10 3.26 98.9 36 3 5 6.20 99.1 Iron4 14 3.95 99.2 37 4 5 5.50 99.2 Iron5 6 4.83 99.0 35 5 3 5.97 99.26 7 5.04 98.9 35 6 5 6.07 99.17 7 5.07 98.6 35 Fouling/Organic 7 4 5.93 99.0 Fouling/Organics

Second Stage Vessel #29 Second Stage Vessel #29

1 7 5.08 98.2 33 1 3 6.06 99.0 Fouling/Organics2 7 5.19 98.7 33 2 5 6.05 99.03 7 5.19 98.6 34 3 5 6.02 99.14 7 5.27 99.2 34 4 5 6.00 99.25 7 5.60 99.1 34 5 5 6.03 99.16 7 5.60 99.0 34 6 5 6.06 99.17 6 5.04 98.6 33 7 5 6.00 98.0

HIGH PH CLEANING TEST

Two lead elements from a first stage vessel and one tail end element from a second stage vessel were removed from train 3 and cleaned with a high pH cleaner.

Results of the test (below) showed no significant improvement in normalized flow or rejection following the high pH clean.

High pH OSCAR Test 7-16-15 (Train-C)

Position Serial # Test Delta psiNormalized Flow

(gpm) Normalized Reject % Notes

1 091021585Pre-Clean 5 5.08 97.4

Extruding VexarPost-Clean 5 6.19 92.4

1 101011381Pre-Clean 5 5.06 98.4

Post-Clean 5 6.07 98.5

14* 90910267Pre-Clean 6 6.05 99.0

Post-Clean 6 6.33 99.1

CLEANING OPTIMIZATION

Since no significant improvement was observed from the high pH cleaning it was not performed on the whole system.

Cleaning procedure was modified to increase recirculation time from 2 to 2.5 hours and the 30 minute soak was eliminated.

This change was implemented in the cleanings of trains 1 and 2 in July 2015.

SPECIFI

C

FLUX

NORMALIZED

DP

STAGE

NORMALIZED

DP

NORMALIZEDPERMEATECODUCTIVITY

R.A.R. EXPANSION

Five new San Diego Formation groundwater wells will be added for a combined capacity of 5 mgd.

Addition of three new RO trains to treat the additional 5 mgd of well water.

Will require expansion of the RO pretreatment system as well as the pipelines.

Existing R.A.R. RO trains will be shutdown for a period of 6-8 months for the completion of construction of the expansion and integration of the new systems with the existing.

PREPARING FOR A LONG-TERM SHUTDOWN

CIP’s for all trains should be performed prior to the shutdown

Elements should then be preserved by filling the pressure vessels with the selected preservative solution.

The solution should be circulated in order minimize the amount of air in the system

Depending on the preservative being used, some may need to be replenished or replaced after a certain period of time.

PRESERVATIVES BEING CONSIDERED

Proprietary Chemicals: Kinglee Technologies: MemStor Avista Technologies: Safeguard 100 Alkema: C-221

Generic Chemicals Sodium Bisulfite Citric Acid

PRESERVATIVE TEST PLAN

Five elements will be removed from the system and cleaned off site.

Those elements will then be wet tested after cleaning

Elements will then be immersed in a barrel containing one of the 5 preservatives being considered.

Elements will remain in the same preservative solution for a period of 6 months.

After 6 months the elements will be removed and once again wet tested and the results will be compared to the wet test data taken prior to preservation.

PRESERVATIVE TEST EVALUATION

Together SPI and Sweetwater staff will evaluate the test data to select a preservative for use during the long-term shutdown.

Criteria to be considered during selection process: Total cost of the preservative

Effect on performance Rejection

Flow

QUESTIONS?