ON-SITEOPTIONSFORTRANSFORMERINSULATIONMANAGEMENT

(With an Historical Perspective)

Andy Bartram I.Eng MIET, Sales Manager

Electrical Oil Services Ltd

ABSTRACT

This paper seeks to review the history of insulating oil management (oil and paper) beginning with an historical perspective, expectations and how oil monitoring and maintenance standards have changed over the years.

Consider how paper insulation determines the ultimate life of the transformer and how the degradation of insulating oil in service reduces the mechanical strength of the paper.

What are the pressures facing todays transformer owners?

Do transformer owners have to take action when oil is “poor” or not?

If so when and what options?

With Electrical Oil Services at the forefront of insulating oil supply and management for well over 60 years the author draws on his and the experience of former colleagues to try and put all this in perspective, ending with some worked examples of in-situ oil regeneration carried out with the intention of improving the transformers health and extending their potential life.

INTRODUCTION

With today’s focus on transformer life extension there has never been a better time to look again at your transformer insulation system.

Having a working knowledge of insulation degradation and options for site treatment can help today’s hard pressed site engineer manage his or her transformers with confidence.

Electrical Oil Services and her predecessor companies have been at the forefront of insulating oil supply, reclamation and on-site treatment for over 50 years. Drawing on this long history of close involvement with the UK electricity supply industry this presentation seeks to review insulating oil degradation (from a non-chemist / site engineers perspective) with reference to BSEN60422:2013 and consider the latest in-situ oil regeneration techniques for “deep cleaning” insulation systems and extending the potential life of transformers.

Some history

In the 1941 edition of the J&P Transformer book[1] the maximum permitted Acid Value for Class A or Class B mineral insulating oil supplied into UK (built) transformers was 0.2mgKOH/gand reflected the limits set out in BS148:1933. Compare this then with the limit given in the current version of IEC 60296:2012 of just 0.01 mgKOH/g [2] and you will begin to get a sense of how our attitude to transformer oil management has changed in the intervening years since the eighth edition of the J&P handbook was published.

Back in the 1930s and continuing until the 1950s a transformer purchaser could specify either Class A orClass B insulating oils for his new transformer (first fill) depending on the likely maximum temperature the oil was ever likely to reach. Class A insulating oil should be used where the maximum oil temperature would exceed 80ºC while class B could be used when “the maximum oil temperature does not exceed 75ºC” [3]

The class of oil chosen would in theory have be critical it reflected the oils Tendency to Deposit Sludge under the then BS test which was carried out at 150ºC for 45 hours, allowable %age sludge after the test was 0.02% for a Class A oil and 0.1% for a Class B oil. By way of comparison the Oxidation Stability test IEC 61125:1992 (Method C)specifies a test duration of 164hours (for uninhibited oils) at 120ºC with a resulting maximum permissible sludge of 0.8%.

In practice transformer users would use the design criteria as set out in British Standard 171 which assumed an ambient temperature of no greater than 35ºC and a top oil temperature of 50ºC – this gives a theoretical maximum oil temperature of 85ºC – many users would still specify a class B oil on the basis that ambient temperature would not be 35ºC and or site loading would not be high enough to give a top oil temperature of 50ºC.

Following privatisation of the United Kingdom Electricity Supply Industry, under pressure from the European Union to further open up UK markets to overseas transformer manufacturers, BS171 was replaced by BS EN 60076 in the early 1990s – a move that at the time was widely seen as a retrograde step for the UK ESI but meant the newly powerful procurement specialists could look overseas for alternative suppliers of transformers.

Along with the liberalisation of the transformer marketthe development of computer aided designmeant that designers could for the first time produce transformers that were fit for purpose andcontained less steel, copper paper and of course, insulating oil.

With lower material costs, the removal of the BS171 “barrier to entry” and arguably at the time (and today?) a price-driven market for new transformers, the scene was set for a new (for the UK market) breed of transformer that we can argue, needed to be looked after, managed and nurtured through life far more than the older CEGB specified transformers ever had to with their built in margin for error, standardized designs, more oil within.

Throughout the 1990’s the UK witnessed a “dash for gas” where previously “rare” natural gas was now allowed to be used as an alternative to coal and oil in conventional power stations. Gas fired power stations sprang up all over the UK with, at the time, a stated expected life span of 15 years. With such a short term investment the majority of the generator transformers specified for these new gas stations were, shall we say “built to a cost” – why design a transformer to last 40+ years when a 15 year life-span was required? In addition these new breed of CCGT power stations were, in the main designed to operate under base load (continuous running) as gas prices were cheap and (allegedly) plentiful.

And so, over 25 years later we bare witness to transformers that were designed to last 15 years, failing, or at the very least showing a degree of insulation wear and tear normally associated with generator transformers twice their age.This degree of insulation degradation is likely due in part to the original design but also, and possibly more significantly, a change in the duty of these transformers – no longer base load operation but on and off load two or three times a day or with long periods of inactivity. Many of these generator transformers, specified for base load operation, were supplied without any forced cooling, relying on the load building up and running at “base load”, steady state for days or weeks at a time. This gives the transformer’s natural cooling sufficient time to “get going” and establish an efficient cooling pattern. Subject the same transformer to a two-shifting regime howeverand owners have found that by the time the transformer’s natural cooling pattern has been fully established it’s time to switch off the transformer, the result is insulation that for frequent periods has seen operating temperatures without any real cooling effect. Paper overheating, embrittlement and failure can result.

What about the utilities?

The historical picture painted above largely focuses on the generation sector of the UK market, with the dash for gas, the emerging dominance of the procurement specialist, fewer (and cheaper) construction materials and, it has to be said, a loss of experience and expertise from the industry combined with a lack of investment and training in young engineers who wish to remain at the operational end of our business and not fast tracked into senior management as is often the case.

The six UK Distribution Network Operators (DNOs) and their 14 separately regulated areas are subject to price control and regulation through the body known as OFGEM (Office for Gas and Electricity Markets).OFGEM regulate the electricity markets on a 5 – 8 year cycle. For the period 1st April 2015 to 31st March 2023 RIIO-ED1[4] (Revenue = Incentives + Innovations + Outputs) dictates the revenue available for investment by the DNOs.

The emphasis is very much on life extension of existing assets over and above outright replacement. When it comes to power transformers with system voltages between 33kV and 132kV DNOs have sought to arrive at a target population that are suitable candidates for investment for life extension – example criteria would be:

  • The percentage life remaining of the paper insulation (based on furan analysis over time)
  • Total Acid Number of the oil
  • Any history of faults
  • DGA history
  • Overall physical condition
  • Availability of spares

It may surprise some but many of the transformers identified as candidates for life extension investment are well over 40 years old, they were, however built to exacting BS171 and BEBST2 standards and, just as importantly, most have operated in parallel with a sister transformer with each transformer taking only half the design load. Paper insulation tensile strength as measured through furan analysis remains good with at least “half life” remaining.

With pressure to make generator transformers last longer than their 15-20 years expected life, older conventional and nuclear stations still needed to “keep the lights on” whilst alternatives can be developed and the pressure on the DNOs from OFGEM to extend the life of existing transformers, there is a growing focus on transformer asset management focusing on the insulation system.

A transformer’s insulation system ultimately defines the life of the transformer (providing it doesn’t die of something else in the meantime!). There is an easy analogy with the human body, if you exercise frequently and regulate what you eat and drink you will on average live longer than someone who doesn’t, the analogy goes further – if you are fit and healthy your body is far more able to withstand illness and disease than if you are not, for body substitute transformer insulation system.

Looking after your paper insulation - water

Let’s get one thing straight – transformers are dried in the factory before dispatch to the end user. Power transformers are typically dried so the cellulose contains less than 0.5% water by dry mass of paper. Once the paper has been impregnated with “dry” insulating oil then this figure may rise to 1% or slightly less.

What does this mean in practice for an in-service transformer?

Assume the following case study

145MVA, 132/145kV

HSPT Generator transformer

1995

Oil volume 34,704 Litres

Measured water content in oil (over time and allowing for stable conditions) 23ppm @50ºC

Assume for a 132kV transformer weight of insulation is 10% weight of the oil

Weight of oil in transformer is 0.87 x 34704 Litres = 30,019 kg

Water in oil can be calculated at 34704 x 23ppm / 106 = 0.80 litres

From a typical paper oil equilibrium graph, figure 1 below 23ppm @ 50ºC equates to 2.3% water in paper

Or

3019kg x 2.3% / 100 = 69.4 litres

When installed and filled with mineral oil with a paper water content of 1% there would be 3019kg x 1% / 100 = 30 litres of water

FIGURE 1

Keeping water in paper insulation to these “as installed” levels means you already are part way to maximizing your transformer’s life. Ignoring leaks, poor maintenance practices, inattention to breathers and overloading transformers for long periods can all lead to water levels increasing either due to atmospheric contamination or cellulose degradation. Drying a “wet” transformer on–site is hardly ever effective unless you are prepared to take some pretty drastic and costly measures involving vacuum/ heat cycles and low frequency short circuit drying usually requiring several return visits over 2-3 years allowing sufficient time for the remaining water in the paper to return to stable / equilibrium conditions with the oil. Once dry, or perhaps from new owners should consider the installation of a proprietary molecular sieve type device to help keep the oil and hence the paper, dry.

Over the years much has been written on the subject of water in paper insulation, not least by former Carless Refining & Marketingemployee and transformer specialist Charles P. Tart in his paper given to the National Grid conference in 1996[5]. “All transformers are “wet”. Wet is a relative term….processing the oil because it is “wet” is not the end of the story. What is being processed is the transformer insulation system, both the liquid and solid parts, and the relationship between these two parts should always be understood before any work is undertaken.What is important is the degree of wetness plus the size and type of the transformer (in 2016 I would add the transformer’s commercial application Auth).

FIGURE 2

Figure 2 above is taken from CEGB transformer oil maintenance documents and remains, to this day a very useful guide to help the practical transformer engineer keep his or her transformer water content to recommended levels. As ever, these graphs assume a) representative samples are being taken by trained operatives and b) the transformer has been on-load (ideally) and operating under steady state load conditions for some weeks ensuring a state of equilibrium between water in the oil and water in the paper.

And acidity?

Now that we no longer have to decide between Class A or Class B oils and their varying tendencies to deposit sludge in your transformer we could hope that oil oxidation leading to acid and sludge formation is a thing of the past – well, yes and no.

Whist there is no doubt that modern mineral insulating oils are carefully refined to ensure top quality and stable performance under exacting conditions it remains a fact that unsuitable specification, poor transformer design, insufficient cooling, poor construction can and do all lead to degradation of insulating systems (oil and paper) and as already discussed it is not uncommon to find relatively young transformers containing highly oxidised insulating oil and low paper strength for one or all of the reasons listed.

Mineral insulating oil is supplied as either uninhibited or inhibited, the latter having artificial inhibitors added to the oil in the refinery to give the oil extended resistance to oxidation due to high operating temperatures and exposure to oxygen in air. Inhibitors can and are often added to in-service transformers, particularly following on-site regeneration processes where the oil is already partly aged.

It should be noted that in reality there is no such thing as an uninhibited mineral insulating oil – a bit of a glib statement but it should be remembered that a mineral insulating oil refined in a modern, hydro-treating refinery will retain sufficient sulphur and aromatic (the smelly part of oil) content to enable the oil to pass the standard 164hr oxidation stability test as set out in IEC 61125:1992 (Method C). Good sulphur and not the corrosive stuff, is a natural anti-oxidant and this is why a good performing uninhibited mineral insulating oil should not be over-refined, over-refining can remove all sulphur and aromatics leading ultimately to a medicinal white oil which is excellent for body oils but not so as a transformer oil unless artificial inhibitors are added – it then becomes an inhibited mineral insulating oil.

During a transformer’s service life heavy acid sludge formation is no longer as common as it was in the 1930s when “new oil” could and was supplied at starting acidities of 0.2mgKOH/g instead of 0.01mgKOH/g (max)[6]specified in today’s standards.

Permanent damage to paper insulation is considered to begin at acid values (TAN) of around 0.08 – 0.1 mgKOH/g with modern transformer asset management practice (based on BSEN 60422:2013) recommending action thereafter[7] depending on the transformers importance to the user. Note there is not always a straightforward correlation between transformer operating voltage and importance to the user, a simple 1MVA 11kV transformer containing 1200 litres of insulating oil could be vital for the production paper in a mill for example and therefore carry a similar importance as a 400kV generator transformer to its’ owners.

FIGURE 3FIGURE 4

Comparison of IFT and AcidityDepletion of “natural” inhibitors

Figures 3 and 4 are useful practical guides to oil and paper ageing and when to intervene, in this case regenerate the insulating oil in such a manner as to “deep clean” the solid, paper insulation.

Both graphs show action points of 0.1mgKOH/g with Figure 3 adding Interfacial Tension (IFT – see below) as an additional metric.

Figure 4 is intended as a graphical representation showing how insulating oil behaves over the years with point A representing the time at which the oils “natural” inhibitors have depleted to such an extent that they can no longer resist the process of oxidation. Point B indicates oil regeneration returning the oil to “as new” but with the addition of an artificial inhibitor to restore the oils oxidation stability.

And other parameters?

Other parameters can be monitored in addition to water and acidity that can serve as a useful guide for the transformer owner when it comes to assessing the overall health of the unit and deciding what action to take, if any, to treat the oil/paper.

  1. Dielectric Dissipation Factor (DDF) has traditionally been an oil test carried out by oil producers and was not necessarily a condition monitoring test, however DDF is now listed in BSEN 60422:2006 with “good” , “fair” and ”poor” action limits depending on the test result.

DDF in simple terms will give a “poorer” result where there are high levels of dissolved polar contaminants present in the oil, these contaminants are often found as a result of oil oxidation and lead to lower insulation resistance and larger transformer losses. DDF is measured under AC conditions as opposed to…..

  1. Resistivity – which examines the oil in a similar fashion but under DC test conditions. This test is an old favourite of the CEGB where it was used as a “catch all” site test, invariably additional tests would have to be carried out where a “poor” result was obtained. DDF is perhaps a more useful test these days and is preferred to resistivity although typical test sets produce both results as a matter of course.
  2. Interfacial Tension (IFT) – has also crept into our consciousness over recent years and, along with DDF is routinely used as a progress check when carrying out on-site oil regeneration, being one of the last parameters to reach acceptable limits.

BSEN 60422:2006 gives a useful summary of this test “The interfacial tension between oil and water provides a means of detecting soluble polar contaminants and products of degradation. This characteristic changes fairly rapidly during the initial stages of ageing but levels off when deterioration is still moderate. A rapid decrease of IFT may also be an indication of compatibility problems between the oil and some transformer materials (varnishes, gaskets), or of an accidental contamination when filling with oil” [8]