Cable Rejuvenation Impact on Loss Factor (tan-Delta)

Dear B.F.,

We have treated a run of feeder cables that spans several miles utilizing Ultrinium™ 733 rejuvenation fluid.  The cables are 35kV class XLPE and have 750 kcmil aluminum conductors.  We have noticed a decrease in voltage drop since the cables were treated.  How, all knowing frog, might rejuvenation affect the power factor and the capacitance of the cable?

Regards,

Ben Miller

The Woodlands, Texas

Dear Ben-

I have been to the site that you describe.  I particularly like the way you laid the cable near the drainage ditch.  I could splash in the water and catch a few morsels to eat throughout my visit.  You might recognize the picture nearby that I snapped of my colleague near the circuit in question.

There has been some work to measure the change in loss factor or tan-delta as silane rejuvenation fluid is absorbed into polyethylene.  The primary purpose of that work was to ascertain if loss-factor measurements could provide any useful diagnostic information after a cable had been treated.  The answer to that question is unequivocally no.  However, the effort could also shed some light on your query.  The work focused specifically on phenylmethyldimethoxysilane monomer (PMDMS), which is the dominant component used in Perficio™ 011 fluid and in CableCURE^®/XL fluid and its predecessor fluid.  The Ultrinium™ fluid you utilized is similar but not precisely the same.  As you will recognize from the balance of the discussion, the differences are of little relevance to the question at hand, although they have profound impacts on post injection reliability.  My colleague, Dr. David Busby, will be presenting a paper on October 18, 2010 at CIGRÉ Canada, titled “Cable Rejuvenation Mechanisms: An Update.”  It explains with a great deal of specificity the differences between the commercially available technologies.

A paper written by my colleague, Glen Bertini, in July 2001, titled “Loss Factor Measurements on Power Cables treated with CableCURE^® Technology” reviewed the work that had been done in this area.  First my old friend Herko Oldehoff of the German utility EWE reported his work in 1994, “Sanierung (Silikonisierung) von 20-kVKunststoffkabeln,” in Elektrizitäts Wirtschaft, (Jg. 93, H. 26. S.1686-1688).  Herr Olderhoff found …

“… that while AC breakdown strength increased after approximately 120 days (after treatment) by a factor of 2.1 on test sample one and by a factor of 1.8 on test sample 2, that the accompanying changes in the loss factor at 0.1 Hz were +5.1% and +341% respectively.  If this utter lack of correlation is not enough for the skeptical reader, it gets worse.  Each of the two tests were samples of the same cable type manufactured in 1982.  The 20kV cables each had 240 mm² conductors (approximately 500 kcm) and each test included samples for each of three phases. One might expect at least some uniformity between the three phases, but in fact there is a wide inter-phase variation, which the average results reported above tend to mask.  For test sample one, the highest post-treatment loss factor is 5.423 and the lowest value is 4.412.  Compared to the average loss factor of 4.755, the range of inter-phase deviation is 21.3%.  For test sample two, the highest post-treatment loss factor is 49.672 and the lowest value is 6.921.  Compared to the average loss factor of 23.513, the range of inter-phase deviation is 182%.  The variability of the inter-phase AC breakdown values was no more than 1X operating voltage (U₀) or 16.7%."

Herr Olderhoff concluded that there is absolutely no relationship between loss factor measurements and actual cable AC breakdown performance.

In short, the loss factor is very variable after treatment with fluid at 0.1 Hz.  The confounding, however, is not limited to 0.1 Hz.  As you can see from the nearby graph, there is a very complex time-dependent and frequency-dependent relationship between treatment with PMDMS and loss factor or tangent delta.  At 60 Hz where your query lies, that loss factor first goes down and then goes up.  After 100 days the loss factor has increased from about 0.0003 to about 0.0008.

Why is this so?  The introduction of PMDMS changes a host of variables in the cable and these variables are all time dependent.  Consider (as Bertini did in the aforementioned paper) that …

• Water is removed from the conductor shield, which changes its electrical properties.
• A variety of chemical species diffuse into the conductor shield and change its electrical properties.  All of these species are dielectric so their adsorption onto carbon black agglomerates generally increases the conductor shield resistivity.
• Chemical reactions and diffusion into and out of the shield change with time.
• Water is removed from the insulation which changes its electrical properties.
• A variety of chemical species diffuse into the insulation and change its electrical properties.  All of these species are dielectric so their absorption into the polymer increases its dielectric strength and increases its permittivity relative to virgin polymer.  Chemical reactions and diffusion into and out of the insulation change with time.  The distribution of these species is not radially uniform.  The radial non-uniformity also changes with time.

There are effectively no changes to the insulation shield, as the concentrations of treatment species are inconsequentially small due to the shield's proximity to the surrounding environment and the thermodynamic sink the soil represents.  Some of the complex changes outlined above would tend to increase the loss factor and others would tend to decrease the loss factor.

Ben, the short answer is:  Injection most certainly does affect the permittivity and therefore the capacitance of the cable.  The effect is time dependent, quite unpredictable, modest, and will dissipate over the life of the treatment.

Missing Texas,

Thermonuclear