Why Do Hospital Generators Keep Failing?
It also happened in Houston during Tropical Storm Allison when hospitals got flooded and found out that their generators didn't work well when they were under water.
Apparently generators failing in hospitals are not as unusual as one might expect.
Uno
They used to test the generator on a 3 monthly schedule. But I think that stopped a couple of years ago.
Another update: Uupgrade work is well underway.
New transformers are being delivered next month. Then in another 4 weeks, they're switching over to the new supply. There will be a 24 hour period with no mains power, and the hospital will be powered by their generator.
The big issue is that a lot of medical equipment has never been tested on the generator, because it had always been assumed that mains supply interruptions would be short. For example, prior to routine generator testing, all big power hogs, like MRI scanners would be shut down and powered off for the duration.
A bunch of people have gone round contacting the various equipment manufacturers asking about generators. The MRI scanner manufacturers have come back saying that the scanners should not be operated on generator power, as the voltage quality is unlikely to be good enough for their sensitive equipment. However, continuous generator power must be provided to the superconductor support equipment.
The CT manufacturer says that the voltage input to the scanner must be within +/- 5% of nominal during a scan. They say the scanner may work if the voltage fluctuates more than this, but the images may be blurred or distorted, or the scanner may abort the scan. The electricians aren't sure that our 700 kVA generator will handle the CT scanner going from 1 kW to 200 kW (it's nominal maximum rating) instantly without the voltage going out of range. They're planning to get a guy with a power analyser in to see what the voltages/currents actually do during "normal" use, and if the numbers look reasonable, they'll test it on the generator
If it doesn't work, then we'll have time to arrange rental of a bigger generator with better voltage regulatotion on a temporary basis - although the main electrician did say that he wasn't sure if any generator supplier could guarantee voltage regulation under "max power" scanning conditions.
If it doesn't work, then we'll have time to arrange rental of a bigger generator with better voltage regulatotion on a temporary basis -
What kind of micky mouse operation is this hospital? Centralized line interactive ups has been the standard for hospital power backup for well over a decade.
are power outages seriously a common thing in the US?
I brought that idea up at my hospital since they got power outages all the time, really short blips but it was enough to cause computers to shut down or freeze, and was probably hard on medical equipment too.
They had some weird political reason as to why they could not do it. It had to be hospital grade or something. I find that's silly, it's best to have a non hospital grade system that can be bypassed, than to not have it at all.
What kind of micky mouse operation is this hospital? Centralized line interactive ups has been the standard for hospital power backup for well over a decade.
wow, one appliance with a 200 kW startup load!!!!!!
Is this a diesel generator? Usually most diesels won't have that much of a problem adjusting rapidly to a huge new load.
A hospital is a big load, and not all equipment is "critical". More importantly, a lot of equipment has very "difficult" load characteristics.I agree, seems odd that there appears to be a complete lack of centralized or even decentralized large UPS/battery capability on critical equipment in a hospital.
Its like I said before, critical systems should have batteries in between them and whatever the power source is. Critical computers should either have an UPS under the desk or should be a laptop in a dock.
A hospital is a big load, and not all equipment is "critical". More importantly, a lot of equipment has very "difficult" load characteristics.
E.g. a surgical laser might be designed for a pulse rate of 10 Hz, so must be able to charge its capacitors is 0.1 s, at a charge current of 100A. But a realistic use, would be 1 Hz, so that every time the laser fires, the lights dim for a fraction of a second as the caps recharge. This type of load is ill suited to UPSs. I know of one hospital that was testing one, and even though the OR suite had a big central UPS, the laser actually caused the UPS to trip-off (it didn't actually cut off the power, but it switched from dual-conversion mode to "fault bypass" mode, where it switched to unprotected mains power having tripped on an "inverter fault").
Similarly, a high end MRI scanner might take 0.5s current pulses of 300A @ 400V to operate its control magnets, but for 90% of the time, the load will be near zero. Again, this sort of load is poorly suited to a UPS, and the manufactures specifically recommended that the MRI scanners NOT go on a UPS, with the exception of the control/data acquisition computer, for reasons of excessive cost, maintenance requirement and space requirement.
The hospital does have a large central UPS for critical areas - the main problem in the original incident was that it drained when the generator failed to start.
Learn something new every day. I will keep my IT experience in the Army, where most of the time our stuff is ruggedized and built for less than ideal conditions. No medical facility work for me anytime soon!
A hospital is a big load, and not all equipment is "critical". More importantly, a lot of equipment has very "difficult" load characteristics.
E.g. a surgical laser might be designed for a pulse rate of 10 Hz, so must be able to charge its capacitors is 0.1 s, at a charge current of 100A. But a realistic use, would be 1 Hz, so that every time the laser fires, the lights dim for a fraction of a second as the caps recharge. This type of load is ill suited to UPSs. I know of one hospital that was testing one, and even though the OR suite had a big central UPS, the laser actually caused the UPS to trip-off (it didn't actually cut off the power, but it switched from dual-conversion mode to "fault bypass" mode, where it switched to unprotected mains power having tripped on an "inverter fault").
Similarly, a high end MRI scanner might take 0.5s current pulses of 300A @ 400V to operate its control magnets, but for 90% of the time, the load will be near zero. Again, this sort of load is poorly suited to a UPS, and the manufactures specifically recommended that the MRI scanners NOT go on a UPS, with the exception of the control/data acquisition computer, for reasons of excessive cost, maintenance requirement and space requirement.
The hospital does have a large central UPS for critical areas - the main problem in the original incident was that it drained when the generator failed to start.
Any consideration to flywheel system to keep things up until Generators kick in? Looks like these Hospitals used Flywheels instead of Battery UPS's to keep there imaging systems up.
http://www.vyconenergy.com/pq/pdfs/130626_VYCON_BannerCaseStudy.pdf
http://www.vyconenergy.com/pq/pdfs/130802_VYCON_TexasScottishRiteCaseStudy.pdf
are power outages seriously a common thing in the US?
It's surprising that hospitals don't have at least two generators for redundancy. Guess they're too cheap. Then something bad does happen and they end up paying more in the long run from all the lawsuits.
There is something seriously wrong with the way the OP's hospital is, and has been, handling emergency power.
Sadly this is "situation normal" for lots of orgs. Generators and their operations are a pretty large line item which makes some finance people target it as "low hanging fruit." I find that most people have serious difficulties evaluating risk which as a combo makes something like "Generator testing and maintenance" something they feel they can red line off.
Then people die. Lawsuits etc. But that is a whole different budget!! amirite? Those lawyers are already on the payroll and they are bored!