1

Transformer Bushing Monitoring - A Case History

George Binas, PSE&G Jeff Benach, On-Line Monitoring Inc.

Abstract

De-regulation of Electric Utility Industry has forced the industry to operate as a “for profit enterprise” in a competitive field with no geographical barriers. In this new environment Safety, System Reliability, Productivity, Cost reduction, and Customer Satisfaction are becoming main drivers for the Utilities. Early retirement invariably leads to loss of experience and knowledge. Also, in an age of cutbacks, scheduled maintenance suffers. Today, many Electric Utilities are stretching their maintenance cycle from every two years, out to five years or more. To meet this challenge without increasing the danger to personnel or reducing the reliability of distribution system, it becomes prudent to monitor the substation HV equipment while on-line. PSE&G in March of 1977 decided to implement a prototype on-service monitoring system program to assess the technical feasibility and cost-effectiveness of this technology.

The paper discusses the use of the SOS Tanδ system in the on-line monitoring of transformer bushings and CTs at PSE&G. A case history is presented illustrating the prevention of a catastrophic failure through the use of the system’s Data, Alarm Reports, Graphs and Resolution functions. Finally a cost/benefit evaluation will be provided in assessing the overall merits of the system. Background

In preparation for the coming economic climate within the Electric Utility industry, it will be more desirable to maintain substation equipment on-service for as long as possible while maximizing its output capacity. Increasing the intervals of periodic testing with existing time-based preventive maintenance programs expands the risk of disruptive failures in substation equipment due to High Voltage (HV) equipment insulation breakdown.

Present insulation testing techniques are periodic, labor intensive, and require the equipment to be taken out of service. Insulation deterioration dur ing intervals between scheduled testing is typically undetected, introducing a significant risk element, particularly in an aging infrastructure. Further, the off-line testing is performed at lower voltages and at no load conditions. The reliability and accuracy of the measurements taken in this manner is questionable given that this technique doesn’t fully duplicate the operational conditions of the HV equipment.

A study completed by the Maintain Electric team at PSE&G found that new on-service Monitoring and testing techniques for HV insulation systems can have an impact on future condition-based maintenance programs. To this end it champions the concept of the On-Line Monitoring Systems (OMS) for transformer bushings and CTs.

The main benefit of these systems will depend on the interest and expertise in studying the data, extracting behavioral patterns, and adopting maintenance policies that will enhance the life and reliability of the substation HV equipment.

After considerable research and discussions with manufacturer representatives and other user utilities, the On-Line Monitoring Inc. SOS Tanδ Continuous Monitoring Sys tem – was selected for field evaluation.

2

Summary of key findings

Information on field installation and operating experience of On-Line Monitoring Inc. SOS Tanδ continuous monitoring system at a PSE&G substation test site is provided here. Particular aspects of system cost are also discussed, with the proposition that the cost per monitoring point will be used as one of the selection criteria when considering the benefits from the installation of OMS systems. Finally, certain conclusions are drawn on the value of the OMS systems and the benefits they provide.

Specifically, it will be argued that the OMS concept has a place in the utility inventory of tools to help meet the challenges of the new operational environment. The assessment and the understanding of what needs to be accomplished and the cost effectiveness of the system should dictate the preference between on-line and off-line testing.

Expenditures for the installation of any OMS system in substation equipment should be based on cost/benefits elements associated with the “BEFORE and AFTER” adoption of the OMS. Such an evaluation needs to be expressed in a format that can be easily calculated, manually or by creating a spreadsheet, and presented to decision-makers. A case-by-case assessment needed on substation’s system / equipment mission criticality and the OMS system’s ability to prevent substation failure to the overall cost for such prevention using the on-line system.

The criticality of any substation equipment resides in two domains. One is associated with its particular operation, its cost, and its safety. The other is related to its assigned mission in the overall distribution network. Therefore an equipment failure can be accounted for as introducing a cost element and a risk element. The cost element summarizes the burden from equipment failure itself while the risk element results from the impact that the equipment failure has on the distribution network (outage time, impact on system reliability, etc.). The SOS Tan δ System – Overview

The SOS Tan δ system estimates deterioration in insulation of the on- line equipment by measuring the relative tan δ value of the monitoring equipment. It calculates the tan δ of a unit as a relative value compared with a reference voltage from another unit that it is grouped with. Since the system uses relative measurements, there is a minimum requirement that at least three (3) units be grouped for each phase of an evaluation. The tan δ is a measure of the dielectric losses, which are caused by capacitive leakage current in the dielectric material used to make a bushing core and are generated within the paper insulation of the core. One of the symptoms of insulation deterioration is its increased sensitivity to changes in temperature and voltage, which manifests itself in the form of relative increased dielectric loss.

Tan d and capacitance are properties of the bushing core design and the dielectric material used in the core as demonstrated in the dielectric loss equation:

Pd = 2 π f C V2 tan d, watts

3

Where f = Applied frequency, Hz C = Capacitance of bushing core (C1), farads V = Applied voltage, rms volts Tan d = Dissipation factor, p. u.

Monitoring and measurement of the tan d gives an indication of the quality of the insulation and its sensitivity to temperature and voltage changes. System Hardware The SOS Tan δ monitoring system was installed at a PSEG switching station and monitors 500kV & 230kV bushings of nine (9) single -phase transformers in addition to 18 CT’s in the 500kV yard. The system collects, evaluates, and displays information from the a total of 36 monitoring points. The system is intended to provide an early warning of bushing/CT insulation breakdown. It provides real time information on the operational condition locally and remotely at PSE&G home office in Newark, NJ through dedicated modem lines. Overall system installation The system consists of the following components: I) The Measurement Tap Unit (MTU). One for each monitoring HV device. II) The Capacitor Divider Unit (CDU). Also one for each monitoring device. III) A Junction Box where cabling form various CDU’s is terminated IV) A “Rendezvous cabinet”, inside the control house, that house the system electronics and where signal from the various monitoring points is terminated through underground

cabling. V) A Computer where the SOS software resides and the condition of the monitoring devices are displayed.

C o n t r o l H o u s e

S e t o f t h r e e C T s -1 1 x

S e t o f t h r e e C T s

S e t o f t h r e e C T s – 3 2 x

S e t o f t h r e e C T s S e t o f t h r e e C T s – 1 2 x

S e t o f T h r e e C T s - 2 1 x

T h r e e S i n g l e p h a s e T r a n s f o r m e r s \ 5 0 0 -1

T h r e e s i n g l e p h a s e T r a n s f o r m e r s / 5 0 0 -2

T h r e e s i n g l e p h a s e T r a n s f o r m e r s / 5 0 0- 3

.

4

Control House Multipair wire

Multipair cable

Multipair cable

Multipair cable s Connecting cable

Connecting cable

Direct burial

in trench

in conduit duct

Schematic No. 1: Typical Installation of the AVO IMS at Dean's 500

Concrete trench

Transformer bushing

Single phase transformer

1st Bank of CT's

2nd Bank of CT's

Computer

Direct burial

Measurement Tap Unit (MTU)

Cap-Divider Unit (CDU) Cap-Divider Unit (CDU)

Single screened twisted pair

Junction Box

Randezvous cabinet

HH-9

1TRX-1

HH-34

HH-34

HH-6

/8/7

/ 2 / 3

Cap-Divider Unit (CDU)

Junction Box

Single screened twisted pair wire

Unit (MTU)

Junction Box

Single screened twisted pair wire

Unit (MTU)

**

** For 2 TRX use HH-46 / 47/ 14 For 3TRX use HH-1/ 2 /3

*

* For 2TRX2 use HH-35 For 3TRX use HH-36

Measurement Tap

Measurement Tap

5

Pictorial view of a bushing and CT with Measurement Tap Units (MTU)

System Software

The system is based on the principle of a conventional Schering bridge under software control to measure the tan δ of the insulation. The system has been optimized using software with minimal hardware. The deterioration of insulation is gauged on its sensitivity to changes in the on-line operating conditions. The versatility of the system installed at substations has allowed for continuous accumulation of data. The SOS Tan δ software optimizes the performance to measuring parameters with existing hardware. Data extracted from measurements are manipulated in matrix operations to present normalized condition values for each unit being monitored.

The signals are sampled and the data processed under software control to calculate the capacitance and dissipation factor changes. Since a second unit of equipment is used as a reference standard instead of a standard capacitor, the values are relative to that unit. This principle is extended to several units within the substation.

With specific configurations, relative measurements eliminate common mode effects, such as ambient temperature, operating voltages and load conditions, and accentuate the differences, such as design and sensitivity to changes in insulation.

For this reason, the system requires a minimum of nine monitoring points that must be on-line. A maximum of 120 points can be monitored in the same substation.

6

An SOS monitor screen is shown above. The monitor screen is displayed on the host PC located in the substation, and gives visual indication of the status of the units being monitored.

7

The graph shown above is a relative Tanδ graph from a PSE&G substation. The Tan δ and relative capacitance values are displayed in the center of the graph, with the scale on the lefty axis. Temperature and humidity are also displayed on this graph, and use the scale on the right y-axis. Temperature is displayed at the top of the graph, and humidity is at the bottom. Reporting

Shown above, is a report sent via FAX to the PSE&G Newark office. This report can also be sent via E-Mail if network access is available. The system can also send a message to a numeric phone pager as well. The reporting and alarm levels are configurable by the user. Monitoring and reporting services may also be contracted.

8

Case Study

This case study is presented to illustrate the analysis and the decision making process undertaken to remove a suspected CT from service at one of PSE&G’s substation using the On-Line Monitoring SOS Tan δ system

In March 1997 the On-Line Monitoring System (SOS) was installed configured to monitor the condition of 500 kV & 230 kV bushings and EHV Current Transformers. It monitors bushings of nine (9) single-phase transformers in addition to 18 CTs. The system collects, evaluates, and displays information from a total of 36 monitoring points. The objective is to provide an early warning of the bushing / CT insulation breakdown.

During the first weeks of March, 2000 the system provided indications that the insulation integrity of one of the monitoring EHV CTs was under stress by progressively exhibiting higher and higher tan d numbers. The screen below portraits the condition of that CT, is #8. Finally on March 14, 2000 the system issued an Alarm condition.

9

This CT was saved by SOS Tan δ Monitoring System

The Alarm was received via FAX at PSE&G. A similar FAX was also sent to the vendor’s FAX server, which also monitors SOS sites. At PSE&G, this development was expected, as the device had been under continuous surveillance for several weeks, as the system had notification that the condition value of the insulation was increasing, indicating that the insulation quality of the CT was slowly degrading.

The graph (shown left) shows that the insulation condition of the CT changed radically on March 3, 2000 and then settled down 24 hours later. This change caused a warning alarm at that time, which cleared when the condition value decreased. On March 4, the condition value again changed causing warning alarms until March 15, when the CT was taken out of service after an on-site inspection. The Oil gauge on the CT was pegged at LOW, indicating that the oil levels were low enough to compromise the insulation quality of the CT. After the appropriate measures were taken and an insulation test was performed the CT was put back in

service.

10

A FAX report from March 14 is shown at the right. This FAX

shows the status of all 36 monitored devices in the substation, and indicates that there is an alarm condition for the 500KV CT, CT #8 with a condition value of 99. This report can also be sent via E-Mail if a network connection is available.

Due to the diligence of the SOS Tan δ system and the foresight of PSE&G, the loss of the CT, a power outage, the possible injury to personnel, and collateral damage were On-Line Monitoringided.

The system indicates status of monitored devices by using a simple green, yellow, red visual indication (shown below.).

System Cost Considerations

As a preface, it should be noted that currently, OMS systems are based on pure data collection philosophy. They require the human intervention and experience (when available) to process the raw data they produce for a decision to reached and appropriate responsive action to be taken. The true benefit of IMS trending will come when it evolves to an “intelligent system” more closely emulating human reasoning with automated crisis-response capabilities. Any IMS strategy adopted at the present time is conducive to the risks and opportunities associated with the early phases of new technology development and its application, but it also offer the opportunity to influence on its evolution.

Any cost/benefit model attempting to establish the true value of any OMS system would thus need to track and follow these unfolding developments to accurately capture the benefits that their adoption can bring to the Distribution Operations. Given the difficulty of accurately building such a long-term model, however, it would be argued here that present justification for such expenditures need to be based on pure merit of gains in maintenance cost and operational efficiency. The issue of safety being so profound can be debated all by itself as a sole justification criterion when appropriate.

11

The On-Line Monitoring System is viewed as an important cost driver holding the promise for equipment maintenance cost reduction, improvement of substation equipment performance and enhancement of safety. COST/BENEFIT CASE ASSESSMENT

Evaluating the cost /benefit in the case history presented here we are starting by calculating the cost per monitoring point and we recognize two cost elements: A. Capital cost a) System cost = $65,000

b) Installation cost = $40,000 ` Total = $105,000 $105,000/36 points = $3,000 per point Assuming six year amortization: $3,000/6 yrs = $500 per point per year

B. Operational costs

A) Estimated Annual Operation cost = $4,275 per year b) Miscellaneous costs (computer/facility & al) = $ 100 per year Total = $4,375/36 = $118.75 per point Grand Total system cost per point = COST OF FAILURE

A CT failure will had cost:

a) CT replacement cost $26,000 b) Labor cost for: i) removal of failed CT, ii) Installation of the new one and iii) testing (for i & ii assume 64 Mhrs, a crew of 4 man for two days)

64MhrsX$50 = $3,200 Testing = $670 Total = $3,870 GRAND Total = $29,870

12

Other ancillary factors that may be consider: Clean-up cost, Collateral Damages, Losses from down time, Inconvenience/PR. CONCLUSION

The conclusion emerging from this case is that if the cost per monitoring point is to be compare vs. the cost of a CT replacement then such a system is very attractive, even when you factor in the cost of that particular monitoring point from the start of the monitoring process. It is reasonable then to justify the cost of all other no event monitoring points as insurance you bought against the non-materialized failures.

On the other hand, if you compare the Cost of failure of $29,870 with the $103,650 total system cost it appears that the system becomes cost effective if it successfully provides warnings resulting in the saving of at least four (4) CTs over the depreciation period.

Taken in consideration that the system also monitors transformer bushings you can extrapolate to a scenario that can include not only the cost of bushing failure but also the added cost of transformer damage. Adding up any cost associated with the type of the outage the failure of the particular equipment can create and you improve the argument for the adoption of an OMS strategy at least for the most critical installations.