Concise description of why elpipes can be both reliable and economical

I came close to a deal with ABB on elpipes in September 2010; I needed 3 ABB vice presidents to say yes, and I got 2; was shot down by Willi Paul (ABB Corporate Research; story details and a picture is in this pdf). Willi admitted that ABB will need something to underground a short bit of 800kV lines in order to be able to build them in Europe or the Eastern US, but he said GIL will be easier to develop for 800 kVDC than elpipes, and the ability to reclose quickly will be important. That motivated me to look deeply into GIL.

I sought out Hermann Koch of Siemens, arguably the leading expert on GIL; I met him for 3 hours at the PES General Assembly in July 2011. Hermann does not think GIL at 800kV is workable in Europe, because the 4kV/mm voltage withstand means that at 800kV, GIL needs to be 1.5 meters in diameter. A 12 GW elpipe can be half as large radially (75 cm) as 800kV GIL. Hermann loved the elpipe idea, and tried to help me get a deal with Siemens. Unfortunately, this effort was contemporaneous with the firing of the general manager of the HVDC unit Wilfried Breuer, and just before he was fired, he said no. He was fired for a cost overrun on the cost of developing offshore HVDC for wind farms.

Elpipes follow a new paradigm. They are mass manufactured cars that snap together quickly, and roll under their own power into a conduit, which is identical in its appearance and installation to a gas pipeline (making no sharp corners, like a train track). Elpipes stand on the shoulders of several well-established industries, including metal forming, polymer fabrication, robotic manufacturing, mine trains (for extracting ore from deep inside the earth), high voltage bushings, and the gas pipeline industry. Installing elpipes inside essentially conventional gas pipelines lowers costs because the technologies used are quite mature. Using conventional installation technology also reduces the uncertainty of installation cost. The segments of an elpipe train snap-fit together, then a solder or liquid metal is injected into the conductive splice, and a pressurized dielectric grease is injected into the insulating portion of the snap fit connector. Said grease can of course be designed to crosslink or polymerize, but for the sake of rapid repairability, I think it is best if the joining grease does not cure. I did a paper for the IEEE 2011 Electrical Insulation Conference with Professor Erling Ildstad of NTNU in Trondheim, Norway showing that a special spirally-wound insulation can greatly improve voltage withstand of the insulation, allowing thinner insulation and therefore improved capacity of an elpipe to shed waste heat through the electrical insulation. That paper also showed that 12 GW elpipe can be buried up to two meters deep and still be able to shed its waste heat to the environment.

One key thing I must challenge is the prevalent view among high voltage experts that it is impossible to get the extremely low flaw rate that is needed to allow an elpipe with 50 splices per km to be reliable. The level of reliability needed (flaws in less than 1 in 10^8 joints) is routinely obtained in aircraft, and also in some makes of cars. Who has ever applied the principles of mass manufacturing to HV splices? I propose to transfer splice quality from being workmanship dependent to being addressable by modern factory automation and QC methods. 

All the joints are made and tested at one end of the line, even if the segment is 2000 km long; this means high tech inspection methods can be deployed. In addition, each segment has a brain, and all the brains are connected via an intranet to a controller. Diagnostic software attempts to predict failure, and the movable nature of the elpipe affords an opportunity at least once a day usually, to take half the elpipe out of service for a few hours to swap out any segment showing signs of impending failure. (When one pole is de-energized for maintenance, the other pole remains active with an emergency near ground potential return; one pole can deliver nearly half the power of the bipole, though with double the resistive loss.)

The elpipe train even allows for upgrading transmission capacity over a year or two, by swapping out elpipe segment modules for higher capacity modules in periods of low transmission demand, without ever having to interrupt regular service.

In order for elpipes to move forward, and I believe this to be a critical step for moving away from fossil fuel dependence, the elpipe concept needs a push from people outside the utility industry. This is for multiple reasons, but partly because the utility industry is so risk averse. This is a BIG STEP in an industry most comfortable with small, evolutionary steps.