PowerSystems World 2001, Chicago – Sept 10-12.
As you know, I attended this annual conference held in conjunction with two power electronics conferences, all under the name of PowerSystems World 2001. Overall, the event was poorly organized and confusing to begin with, however, I’ve learned that a new company is taking over conference operations, so things may improve. More significant, the news on September 11 broke just as the conference was getting underway and had everyone completely distracted, so little was accomplished.
There were a few nuggets to extract, however. The plenary session papers were interesting, and I delved into one or two other issues with company reps on duty at their exhibit booths.
The conference proceedings have been published in a 5 lb. 2″ thick hardcover volume, can be ordered for $150 (plus s/h) from Linda McCay, 805-389-6600, firstname.lastname@example.org
Through December, the conference website will provide the agenda and other conference information: http://www.powersystems.com/
The opening plenary session papers were not provided in the proceedings:
“Power Electronics in Power Systems: Technology and Business”
– Vlatko Vlatkovic, General Electric Corporate R&D
With advent of distributed generation, any source that isn’t constant speed synchronized will require significant power conditioning, i.e. “four-leg” converters for fuel cells and microturbines. Such power conversion technology is derived from motor drives using IGBTs (in the 20 KVA to 2.5 MVA range). Large DC systems (100+ MW), e.g., high temperature fuel cells (solid oxide), need conversion from low voltage DC to high voltage AC. Market drivers are strong, but complicated. With utilities’ conservatism, it will be hard to make money, and the technology is changing rapidly. Regulation of transmission systems (and the coming RTOs) means limits on ROI.
The Motor Drive Revolution in the Energy Starved Generation
– Dr. Alexander Lidow, International Rectifier
IR sees huge opportunity, and high payback from power management technology. 57% of electricity consumption goes to motors, most of which are controlled on/off, which is very inefficient. Savings would be substantial if variable speed drives were used in refrigerators, washing machines and air conditioners. Barriers include cost, time-to-market, and the need for multiple engineering skills to apply technologies in analog/digital (A/D), sensors, power semiconductors, and software control. IR has developed “Accelerator” architecture development system and a “ChipSet” to simplify and facilitate the design process.
Tech info: http://www.irf.com/prmU7avc/product-info/motor/
For papers: http://www.irf.com/OxJB7a4c/e/powersys01.htm
Data Center/Facility Infrastructure Design For The Next Millennium
– Neil Rasmussen, American Power Conversion
This paper challenges a number of assumptions about the future power needs of Data Centers. (I have the complete text, which I can forward on request.)
First, an enumeration of the problems driving change:
– Data Center owners get poor ROI on power systems, because they install much more capacity than they can use, and it sits idle, while capital and service costs continue.
– Need to standardize and “drive out variability”. The industry is full of mythology, and there are limited opportunities for systematic learning. Troubling events are random and rare. Installations are one-off custom designs. (reminiscent of nuclear power plants?)
– Can’t predict system requirements. Computer systems become outmoded and are replaced in 2 years, while power systems expected to last for 15 years. That’s 7 generations ahead of the loads being served.
– Systems can’t adapt to increases or decreases, so installations are grossly oversized.
Then, some predictions:
1. No more raised floors – wiring will be overhead, floor will be a slab — cheaper, more secure, air treatment easier to manage, etc.
2. AC will prevail. Need open architecture, so DC will remain limited to inside of closed systems (boxes). Also, adding a new DC drop is expensive.
3. Overloading will occur because power draw of new data equipment varies (2-5 to one) depending on computing activity level, unlike most present-day systems.
4. Instead of focus on total facility loads, discussion will move to “watts/rack”, not “watts per sq.ft”.
5. Modular scalable systems will replace highly (site-specific) engineered solutions. The whole system will be made with cookie-cutter prefabricated modular scalable systems, designed and delivered “just-in-time”.
6. Fast-cycle centers. It now takes 9-18 months to design, purchase, install, and start-up a center. In 2 years, this will drop to 3-6 weeks. This will be accompanied by big changes in the financing and capital structure.
— In another paper from APC, Chris Thompson outlines overall design issues for data centers, pointing out that redundancy is often put in the wrong place, i.e. at the last step before the load. The lack of training for facility personnel means that human factors play a bigger role than they should.
Power Management -Not an Option
– David Kreiss, Kreiss Johnson Technologies
For most industrial and commercial firms, power is a sizable part of their cost of operations, but many have yet to realize that it can and should be managed. The average CFO sees electricity in particular as the least controllable of all – a fixed cost. This view may have been valid in the past, but there have been big changes in the business and power environments, and now, power procurement is negotiated, in terms of both cost and quality. Savings go straight to the bottom line.
Over 100 technical papers, published in the proceedings, were organized under these headings:
-PQ Solutions, -PQ Battery, -PQ Standards, -Distributed Generation, -PQ Monitoring, -PQ Distribution, -PQ Harmonic, -PQ Flicker, -PQ UPS, -PQ TVSS
(The complete program is available as a download:
“PQ Impacts of Distributed Generation”, Roger Dugan, Electrotek Concepts, (page 190) admitting a “pro-utility” bias, went into some detail on complex technical issues arising from the placement of DG on a utility distribution system:
– Fault overcurrent in radial feeders can come from 2 directions; DG must also disconnect.
– Low voltage after interruption- DG needed for voltage support, but can’t come back on until the voltage is restored.
– Reclosing, because most faults are temporary, usually done in less than 1 second (many utilities do it “instantly”). If DG doesn’t do the same, fault won’t clear.
– Transformer connection issues-Wye delta seen as best, but utilities forbid DG to use it, because of grounding dangers.
Engineering solutions can generally be found for these and other technical problems, but there are no simple answers, and there are questions as to who should pay. DG can be used with “no changes” if less than 15% of a feeder load (5-10% if rural).
There were several new technologies for UPS ride-through (i.e., to a standby generator):
– Active Power is developing an Integrated Flywheel Microturbine. The flywheel spins on the same shaft as the turbine rotor, and can provide instant power until fuel is supplied to the microturbine, which then takes over for as long as necessary.
– Precise Power’s written pole technology has been used in a flywheel motor/generator for over 10 years, for short term ride through. Can be integrated with an IC engine for long term backup.
– Metallic Power’s Zinc-Air system is being applied in a rack mounted version for use in data centers. Longer ride through (many hours) at the rack itself avoids reliablity problems of facility based UPS systems (and the siting and emissions issues for gensets).
Some interesting controversies are evident in the world of lead acid battery systems, regarding monitoring and charge balancing.
Charge balancing, or “equalization”– we’ve seen AutoCap previously (UFTO Note – Travel Reports, 29 Oct 2000) proposing a system to float-charge cells individually–don’t know how they’re doing. PowerDesigners is now promoting its “PowerCheq” modules which interconnect adjacent cells in a string, and uses stronger batteries to top off weaker ones. (www.powerdesigners.com) The problem is that greybeards in the industry insist this is the worse thing you can do! In particular, it will mask a problem with a bad cell until the whole line goes down.
Similar reactions are heard to a novel pulse conditioning method, called ReNew-IT, invented by Pulse Tech Products. They apply a unique waveform pulse train to the battery string, explaining that it clears away sulfation deposits that block plates and decrease life. Apparently the military has taken a hearty interest, but another industry greybeard told me–one should just never never apply AC of any kind to a battery.
Battery monitoring also has ideological splits, e.g. whether resistance, impedance, or conductance is the right thing to measure. Btech (www.btechinc.com) says they’re the oldest and best supplier of “battery validation systems” and insist that impedance is the way to go. Their counterparts at Alber say resistance (www.alber.com — they sponsor an annual conference on the subject). (Both show a long list of prominent clients, with a lot of overlap, e.g. the NY Stock Exchange!) And newer arrivals, Midtronic and Vanner (www.vanner.com) have their views as well.
Serveron, a new entrant, is getting an excellent response from major customers with its CellSense technology, which answers all the greybeard concerns and then some. (See 14 May 2001 UFTO Note – On-Line Transformer and Battery Monitoring). CellSense monitors provide continuous measurements of all key physical and electrical parameters needed to characterize the condition of all individual cells as well as the battery system as a whole. www.serveron.com
(They’ll be showing at the IEEE T&D Expo in Atlanta, which opens Oct. 29. I’ll be there as well)