Laszlo, I appreciate your read of the article, and I offer my explanations in italics
Cost?
Less for the MV system based on the copper alone
Component count?
reduced with the MV system, likely losing a whole level of transformation
Higher transients introduced with the vacuum breakers resulting in higher BIL requirements?
Design/mfg detail. We have 4160V systems working today, and without looking at a spec sheet, I don't know where the BIL ended up, but they're out there.
The article's argument about smaller footprints reduces the avilability to certain manufacturers, therefore the pricing reduction claim is dubious.
Agreed that equipment supply is currently limited, but we know of at least two manufacturers to provide a level of competition, and we know of future MV equipment releases from other big players like Square D. Much of the marvel for me with these systems is the reliability of the MV linear-actuator breakers, with lifecycles in excess of 100,000 operations
Eventually you have to step down to a lower voltage to utilize it.
As the article states, 4160V to the UPS; 4160V from the generator. We then envision distribution through Power Distribution Units (PDU) on the order of 500 to 1000 kVA, with 400/230 V distribution to rack systems utilizing the higher power supply voltage
The article presents a single case without the price comparison, it ignores more economic alternatives available for the 'opposing' case.
The components of this 5kV system are common in the larger data centers that we design. It won't fit everything, but show me a 480V system and I'll propose an alternative 5kV system. It really comes down to your end-use equipment. In our systems it's static and rotary UPS systems, and that equipment is available in 4160V.
Remote operators make the safety argument trivial.
Trivial?? What if you need to work on an energized bus? If it's over 40 cal/cm^, you can't legally do it.
I have an Arc-Flash/Incident Energy (IE) appreciation for Electric Rooms like I've never had before. I see from the IEEE 1584 Calcs that there is often an Arc-Flash boundary of 30+ feet for large 480V systems, say 4000A, which are very common in industry. My enforcement of the codes that regulate the handling of Arc-Flash says that I can't even go in those Electric Rooms when the equipment is being interacted with. If it's a 5kV system, that Incident Energy monster lessens significantly, by an order of magnitude.
As I stated many times elsewhere, the science and formulas behind the arc-flash energy so far have proven to be faulty as actual test cases fail to produce the predicted energy levels in actual configurations that match field conditions.[/QUOTE]
IEEE 1584 is essentially the law today. It's what an engineer uses to determine the IE levels. I have no higher or alternative authority to go to. (there are other approved methods, but 1584 is pretty much the industry standard)
I know there are variables that skew the real-world conditions, but the IEEE 1584 committee meets again soon, and whatever they come up with is then going to be my design standard, but until then.... IEEE 1584 2004 is what we design to, and our comparison of comparable LV vs MV systems illustrates that calculation.
John M
