Keeping your fleet in top condition

How long will our trans­form­ers last? This ques­tion is faced by more and more grid oper­a­tors around the world whose fleets are becom­ing increas­ing­ly dat­ed. Pre­vi­ous­ly, the deci­sion to replace a trans­former was often exclu­sive­ly based on its age, regard­less of its actu­al con­di­tion. With a new method, this will all change.

Jan Patrick Linossier and Rolf Funk have racked their brains over the so-called „bath­tub curve“many times already. Both are respon­si­ble for strate­gic asset man­age­ment at the „Rheinis­che NET­ZGe­sellschaft“ (RNG), Rhineland‘s elec­tric­i­ty grid oper­a­tor, and are there­fore tasked with assess­ing the con­di­tion and risk fac­tors of the equip­ment in use. This includes around 100 pow­er trans­form­ers which the dis­tri­b­u­tion sys­tem oper­a­tor man­ages from Cologne for an elec­tric­i­ty grid span­ning rough­ly 20,000 kilo­me­ters. „We have to know the exact con­di­tion of our trans­form­ers. Whether they‘re still ser­vice­able or whether we need to invest in new ones is a ques­tion we are con­stant­ly faced with. The prob­lem is, nobody in the indus­try knows exact­ly how long a trans­former will ulti­mate­ly last,“ explains Linossier, who heads up strate­gic asset man­age­ment at RNG.


The bath­tub curve is a mod­el for describ­ing the sus­cep­ti­bil­i­ty of equip­ment to fail­ure over its entire ser­vice life. Just after com­mis­sion­ing, the fail­ure rate is very high. Once the ini­tial prob­lems have been fixed, the fail­ure rate drops rapid­ly and hov­ers at a con­stant­ly low rate. For a long peri­od of time, this doesn‘t change. But as the sys­tem gets old­er, the prob­lems start to stack up again after a cer­tain point. If you rep­re­sent­ed this trend as a graph, with time along the X‑axis and the fail­ure rate along the Y‑axis, the curve would look some­thing like the cross-sec­tion of a bath­tub.

Jan Patrick Linossier, Head of Asset Man­age­ment at RNG (right), and Alex­ei Babiz­ki, Port­fo­lio Man­ag­er at MR, are hap­py with the results of the pilot study. (© Dirk Moll)

But what Linossier and Funk have been rack­ing their brains over is the fol­low­ing: At what point of the bath­tub curve are their trans­form­ers cur­rent­ly at? When does the fail­ure rate start to climb again? „That is the point at which we should be arrang­ing a replace­ment,“ says Mr. Funk, who spe­cial­izes in strate­gic asset man­age­ment for pow­er trans­form­ers. Since the aver­age age of trans­form­ers at RNG is 45.7 years, this ques­tion is becom­ing increas­ing­ly press­ing. Most of these trans­form­ers were com­mis­sioned back in the 1950s and 1960s. The deci­sion to replace these trans­form­ers not only has tech­ni­cal ram­i­fi­ca­tions, but also has finan­cial con­sid­er­a­tions as well, since a new pow­er trans­former of this scale would cost almost half a mil­lion euros. „Pre­vi­ous­ly we would sim­ply replace the trans­form­ers with new ones after a cer­tain amount of time, but the cost pres­sures today are much high­er. For this rea­son, we want to get the most out of the equip­ment for as long as pos­si­ble,“ says Mr. Linossier.

„There are trans­form­ers used in indus­tri­al fields which have to replaced after 15 years of usage, and then there are those which only oper­ate at 30 per­cent load and have been in oper­a­tion for over 50 years already. So the transformer‘s age alone doesn‘t nec­es­sar­i­ly say any­thing about its con­di­tion.“
Alex­ei Babiz­ki

And RNG isn‘t the only oper­a­tor with a fleet of aging trans­form­ers: In the two decades fol­low­ing the Sec­ond World War, the demand for more and more elec­tric­i­ty increased rapid­ly. New pow­er plants were built, cables were laid, and numer­ous pri­ma­ry sub­sta­tions were erect­ed. Huge swaths of the Ger­man elec­tric­i­ty grid today are still based on the plans drawn up dur­ing this time. And this was not lim­it­ed to Ger­many, but was rather a phe­nom­e­non of all mod­ern indus­tri­al nations. That said, the net­work load fore­casts at the time were much dif­fer­ent.

„With the increas­ing use of renew­able ener­gies, grid struc­tures need to be com­plete­ly redesigned and it‘s there­fore very impor­tant that we know the exact con­di­tion of our trans­form­ers,“ stress­es Mr. Linossier. On the one hand, because the uti­liza­tion of the trans­form­ers changes over time and, on the oth­er, because upgrad­ing the net­work will make some instal­la­tions redun­dant in the future. „If we replace the trans­form­ers too ear­ly, that could turn out lat­er to be a bad invest­ment,“ adds Mr. Funk. At present there is no method for assess­ing the exact con­di­tion of the trans­former fleet.


How­ev­er, all this is set to change thanks to a chance encounter in a tram in Lyon. Alex­ei Babiz­ki, Port­fo­lio Man­ag­er at MR and Markus Zdrallek, a pro­fes­sor at the Uni­ver­si­ty of Wup­per­tal, got into a chance con­ver­sa­tion on the way to the air­port, both hav­ing just attend­ed the same con­gress. Zdrallek holds the senior pro­fes­sor­ship for elec­tri­cal ener­gy sup­ply tech­nol­o­gy and has been work­ing on the con­di­tion assess­ment of equip­ment in elec­tric­i­ty grids for many years. „We want to offer grid oper­a­tors a sci­en­tif­ic basis for intel­li­gent net­work renew­al strate­gies,“ says Mr. Zdrallek, „but pow­er trans­form­ers have not yet been exam­ined in my research.“


The con­di­tion of the trans­form­ers is exam­ined in up to three stages, which can reduce the required down­time of the trans­form­ers:

1 In the first stage, the oper­a­tor pro­vides the data which is already avail­able. This includes, for exam­ple, the age of the trans­form­ers, mea­sure­ments that have already been car­ried out, and main­te­nance his­to­ry infor­ma­tion.

2 The sec­ond stage involves an on-site appraisal which is car­ried out while the trans­form­ers are still in oper­a­tion. In the visu­al inspec­tion, an expert uses a check­list to exam­ine the trans­former based on its exter­nal char­ac­ter­is­tics: Are there any signs of cor­ro­sion or leaks? Is there vis­i­ble dam­age to the bush­ings or fans? How old is the motor-dri­ve unit? For the mea­sure­ments, ther­mo­graph­ic tests are per­formed and oil sam­ples are tak­en.

3 If, after this stage, there are any indi­ca­tions of more sig­nif­i­cant dam­age, the third exam­i­na­tion stage will fol­low. In this stage,exten­sive mea­sure­ments are per­formed, includ­ing a dynam­ic resis­tance mea­sure­ment and par­tial dis­charge mea­sure­ment. How­ev­er, to car­ry out these tests, the trans­former must be removed from the grid.

The data col­lect­ed using this method is eval­u­at­ed. The analy­sis is car­ried out using the algo­rithm devel­oped at the Uni­ver­si­ty of Wup­per­tal in asso­ci­a­tion with MR. Then the data is trans­ferred into two index­es, with one reveal­ing infor­ma­tion about life-time wear and the oth­er about the fail­ure rate. For the assess­ment, the results of the two index­es are trans­ferred into one graph­ic. The con­di­tion of the trans­former fleet can thus be visu­al­ized sim­ply; the more trans­form­ers locat­ed in the quad­rants near the ori­gin, the bet­ter the con­di­tion of the fleet and vice ver­sa.

This imme­di­ate­ly aroused the inter­est of Mr. Babiz­ki, since he was look­ing for a sys­tem for deter­min­ing the con­di­tion of pow­er trans­form­ers. He also knows that this helps pro­vide an impor­tant ser­vice for MR‘s cus­tomers: „It‘s impos­si­ble to give a gen­er­al answer to the ques­tion of how long a trans­former will last. There are trans­form­ers used in indus­tri­al fields which have to replaced after 15 years of usage, and then there are those which only oper­ate at 30 per­cent load and have been in oper­a­tion for over 50 years already. So the transformer‘s age alone doesn‘t nec­es­sar­i­ly say any­thing about its con­di­tion,“ explains Mr. Babiz­ki. He and Mr. Zdrallek quick­ly agreed that this would be the ide­al top­ic for col­lab­o­ra­tion.

„It is of strate­gic impor­tance for us that we know the exact con­di­tion of our trans­form­ers.“Jan Patrick Linossier

„We didn‘t have the well-ground­ed under­stand­ing of trans­form­ers we need­ed, but we did have exten­sive exper­tise in the devel­op­ment of con­di­tion assess­ments,“ recalled Mr. Zdrallek. The researcher hopes that the project will shed more light on how trans­form­ers dete­ri­o­rate over time, since he is also inter­est­ed in the bath­tub curve. „The curve is often described in the the­o­ret­i­cal lit­er­a­ture and has even been proven for some ele­ments of elec­tron­ics. But I was skep­ti­cal of whether the curve even exists for elec­tri­cal grid com­po­nents, and more specif­i­cal­ly for trans­form­ers, because nobody has yet man­aged to prove this,“ adds Mr. Zdrallek. And thus began the col­lab­o­ra­tion between the two part­ners. In a series of meet­ings and work­shops, they worked togeth­er to devel­op an assess­ment sys­tem.


What was still lack­ing, how­ev­er, were trans­form­ers on which they could test their assess­ment sys­tem. Mr. Zdrallek thought imme­di­ate­ly of RNG, with which he had already com­plet­ed numer­ous projects. Linossier and Funk didn‘t have to think about it for long: „The approach seemed very promis­ing to us, so we imme­di­ate­ly agreed to col­lab­o­rate in a pilot study,“ says Mr. Linossier. What made this approach spe­cial is that the sys­tem (see box below) regards the trans­former from two per­spec­tives. On the one hand, there is the long-term view which focus­es on longevi­ty and is thus impor­tant for invest­ment deci­sions. And on the oth­er hand, there is the short-term view which is tar­get­ed at the risk of fail­ure, and influ­ences deci­sion-mak­ing relat­ing to main­te­nance.

Markus Zdrallek, pro­fes­sor at the Uni­ver­si­ty of Wup­per­tal, hopes that the sys­tem will open up new avenues of research for the effects of aging on pow­er trans­form­ers. (© Dirk Moll)

All the main com­po­nents of the trans­form­ers are exam­ined care­ful­ly, from the on-load tap-chang­er and motor-dri­ve unit to the cool­ing fans. A total of 200 para­me­ters ulti­mate­ly influ­ence the out­come of the assess­ment, includ­ing main­te­nance his­to­ry data and results from on-site mea­sure­ments. The col­lect­ed data is then ana­lyzed using a spe­cial algo­rithm. „There are many indi­ca­tors we can use to ascer­tain the con­di­tion of the trans­former. If, for instance, the ther­mo­graph­ic mea­sure­ment shows that a par­tic­u­lar area of the trans­former is hot, it is pos­si­ble to local­ize the fault. Then detailed mea­sure­ment meth­ods are required,“ explains Mr. Zdrallek. Anoth­er cru­cial ele­ment for the assess­ment is the oil analy­sis, which can be car­ried out by MR in the lab­o­ra­to­ry of its sub­sidiary Messko. As Mr. Babiz­ki makes clear, „it‘s a bit like doing someone‘s blood-work at the lab, where an increased white blood cell count might indi­cate an infec­tion, for exam­ple. Sim­i­lar­ly, for trans­former oil, a high­er acetyl val­ue is an indi­ca­tion of elec­tri­cal arc­ing, which may be a sign that the trans­former is in poor con­di­tion.“


RNG pro­vid­ed nine trans­form­ers for the pilot study. „Our choice of trans­form­ers reflect­ed a rep­re­sen­ta­tive sam­ple of our fleet,“ explains Mr. Linossier. After exten­sive tests and sub­se­quent analy­sis, the results were in: All trans­form­ers were found to be in good con­di­tion. For some units, the risk of fail­ure could even be reduced with sim­ple mea­sures, thus extend­ing their ser­vice life. „That‘s the advan­tage of the method we have devel­oped. Not only do we offer our cus­tomers a con­di­tion assess­ment, but also rec­om­men­da­tions for cor­rec­tive action,“ empha­sizes Mr. Babiz­ki.

The part­ners also take spe­cial cas­es into con­sid­er­a­tion. For exam­ple, in one trans­former a high acetyl val­ue sug­gest­ed there was a prob­lem. As Mr. Linossier stress­es, „MR point­ed out that this val­ue can vary depend­ing on mod­el.“ The lim­it-val­ue being exceed­ed in the trans­former oil was there­fore triv­ial, which made renew­al unnec­es­sary. „We designed the algo­rithm to take into account the type of tap chang­er used,“ adds Mr. Babiz­ki.

Mr. Zdrallek is also hap­py with the new method: „It is excit­ing to me from a sci­en­tif­ic point of view. If many net­work oper­a­tors start work­ing with it, then we might be able to use the mass of col­lect­ed data one day to find out whether the bath­tub curve is actu­al­ly real.“ In any case, RNG has decid­ed to use this method to exam­ine the major­i­ty of its fleet of trans­form­ers. „We now have an instru­ment for objec­tive­ly eval­u­at­ing the con­di­tion of our fleet. This is an enor­mous­ly im por­tant devel­op­ment for our strate­gic deci­sion-mak­ing,“ under­scores Mr. Linossier.

Learn more about this collaboration in an interview with Jan Patrick Linossier and Markus Zdrallek.


Would you also like to apply the new method to assess your fleet of trans­form­ers?

Alex­ei Babiz­ki is avail­able for your ques­tions:

Share with your network!

Never miss an issue again!

Click here to subscribe for free.