Backup Power

First step is to list all loads intended to run from inverter or UPS. Doesn't matter at this point if you use watts or amps. Presumably all of these loads use the same nominal supply voltage. If your loads fluctuate or cycle then derive or estimate some kind of mean average.

Now convert al loads into same units either watts or amps. Doesn't matter which, so long as all loads are in same units.

Watts = current in amps multiplied by volts.

Amps = watts divided by volts.

Totalize all loads into a single number for amps. This is your load current.

Determine what the conversion losses of the inverter or UPS is.

Some arun about 10% better ones run as good as 97%

Calculate the theoretical current on battery side. Wattage out must equal wattage input. Add the conversion loss. This is the actual load current placed on the battery bank.

From the battery manufacturer, determine the Amp hour capacity as well as the derating based on Peukerts exponent. This exponent is a function of load curent placed on a battery. All batteries are rated according to a industry wide formula either a ten hour ratign or a twenty hour rating. In the real world you seldom fint the exact same load curent. A scientist named Peukert established the formula for calculating actual run time when load current is greater than this industry rating. The more curretn the less run time you get as a percentage. It is not a linear relationship. If you double the current you end up with less than half the capacity.

This is rather too complex for a simplified answer in one or two paragraphs on a forum. Be aware that several companies now make "battery capacity meters" that calculates actual capacity remaining based on a running calculation using Peukert's formula.

I would add that you want to use "deep discharge cycle" batteries, not common car batteries, unless you are cobbling something together out of parts you already have. You also do not want to run your batteries down much below 50% of their rating if you want them to last and you use them somewhat often. If you are only concerned with rare "emergencies" then you might go as low as 33% or so.

There is a host of related information on the web under topics related to operating a home off the grid, since windmills, solar, and intermittent generator usage all require battery /inverter systems to store power.

Steve , do the UPS companies even recommend anything BUT deep cycle batteries. I would have thought that was a given.

The big difference between what I use and the UPS industry seem to prefer is they use GEL batteries and these are limited to a lower discharge rate than the kind of peak surges my system are subjected to. I usually have to plan on sustained 10 - 12kW surges sustained for as much as 30 minutes on end. That really plays hell with gel batteries. AGM seem to fare much better in that regard.

Sizing UPS systems can be very tricky considering power versus apparent power, etc. Rather than me messing it up and confusing you, I recommend you visit the APC web site (http://www.apc.com) - pick a country, click on "Support", then go to the "Self-Help" section (or just go to http://www.apc.com/support/index.cfm). There are quite a few manuals and white papers, a lot by Neil Rasmussen who is pretty knowledgeable regarding UPS material.

I don't try to push APC products (I use a lot of Tripp-Lite gear myself) but they have a pretty helpful site - their gear is good as well.

Sleddriver, a general question on IT center power versus powering industrial or domestic installations from battery power. Isn't the IT center loading much more steady and thus predicitable. From what I recall back some twenty years ago the computers have a more predictable power demand and under emergency power conditions does the IT operator start up some new sequence that has a high power demand factor?

You've got some pretty good advice above.

The only thing I would add is:

The batteries are sized for the expected average load.

The Inverter must be sized for the maximum expected instantaneous load.

This becomes a factor if there are high peak items such as laser printers

or motors involved.

Good luck,

Pragmatist.

Pragmatist wrote: "The Inverter must be sized for the maximum expected instantaneous load".

Actually this is not true. In fact the big advantage over conventional rotary gensets is the fact inverters typically can handle surge loads of 2X - 3X their normal continuous rating.

I routinely handle 10kW peak demand loads using 6kW inverters systems. An assumption is made here that the battery ampacity, internal resistance, cable lenght and diameter are sized to match the expected loading. Increasing cable lenght between battery and inverter or reducing wire diameter to save money is guaranteed to cripple the system' ability to deliver high peak load current for even short durations.

In most installations the actual DC cable size between battery and inverter is going to be much larger than what would otherwise be indicateed using cable ampacity tables. If possible select cables to keep maximum voltage drop to less than1.5% under peak loads.