Is it feasible to run small office on generator?greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread
How feasible is it to run a small electronic publishing firm on a generator? The office consists of 6-7 computer work stations and two servers. How big a generator? How much gasoline?
-- Libby Alexander (Libbyalex@aol.com), December 22, 1998
Yes. As long as you have fuel. In the S.F. Blackout many such organizations, large and small, ran on generator power. Think I remember numbers of 24 hours to three days being the limit.
-- Diane J. Squire (firstname.lastname@example.org), December 22, 1998.
Let's see, 6 or 7 workstations. Each of these would run 350, perhaps 500 watts, as would the servers. So, maybe you'd need 4500 watts or 5000 watts to power the machines. Add another couple of kW for lights, etc., and you're looking at an 8 to 10 kW generator. [These are rough numbers......can you refine the workstation and server power demands? If so, perhaps you can get by with less.]
These aren't the most fuel efficient things in the world if you use gasoline. Maybe you'll get about 1 to 1 1/2 hour per gallon of gas. (That slurping sound you heard was your tank running dry.)
Options to look at:
Gas or Diesel? Diesel is far less hazardous, a diesel generator is more reliable (and costs more)
Generator only, or generator plus battery pack/inverter? In order to drive the stations you'd probably need as many as 3 inverters and your battery pack should be fairly large [perhaps 16 to 20 T-105s], but you could cut down on the generator size, and on fuel consumption (both from smaller generator and because you wouldn't be running the generator for 8 hours a day).
Can be done. Need better data on the power per station.
-- rocky (email@example.com), December 22, 1998.
For heavens sake, if you do this, get some UPS power supplies. They will cut the surges and power spikes and electrical noise from the generator to something the computers filter circuits can handle. Four good ones would probably be enough.
-- Paul Davis (firstname.lastname@example.org), December 23, 1998.
One thing you can do to lower your power consumption is get a monitor switch and you'll only have to use one monitor for both servers. Leave the keyboards and mice out of it....you can still hook the spare monitor back up later.
Turn the monitors off when they're not being actively used, and plan to backup every open document frequently to prevent data loss. If you can concentrate on one of your servers to serve as both a data and print server by hanging a printer of one of its ports, and back up workstation files to that server, in case you have to shut down workstations, you can still print from the server.
-- Karen Cook (email@example.com), December 23, 1998.
Let's see here... you want to run your firm on generator power for an undetermined period of time... no lights? no fax machine? no copiers? no printers? What about heat? You expect the phones/pbx to be ok? Water still flowing/flushing?
-- Elbow Grease (Elbow_Grease@AutoShop.com), December 23, 1998.
It is said that way more problems to computers are caused by dirty power than by surges.
When you have dirty power (which is likely, if we're going to have rolling blackouts or cascading outages), garden variety surge protectors won't cut it. Even some cheap UPS won't either, if they don't have line conditioning to protect against power sags.
After the Northridge quake, 4 out of 5 of our computers were toast within a few months. (many power anomalies in the area, 4 miles from the epicenter) This happened despite no detectable *surges*, and the computers had been plugged into "surge protectors".
Once I got the UPS with power monitoring software I could see the power problems! Many sags and occasional spikes. Have had not a single problem after putting the UPS in. Actually, only one computer needed a UPS, the others are plugged into line conditioners (about a $70-$100 item).
This is just my experience and semi-educated take from studying up on this. I've heard that some computer power supplies now have their own line conditioning. Don't know. At this point, I'd rather be safe than sorry.
Pray that this is a problem for you. (vs. the alternative) :)
-- D B Spence (firstname.lastname@example.org), December 29, 1998.