# What's wrong with my UPS?

We have a battery & UPS setup on communication systems for an isolated site (wind farm). Last night was the first time the system was put to the test when our grid connection was lost at 11.30 pm, at 8.00 am we had no internet connection, we cannot say precisely at what time the UPS was drained.

The grid was reconnected at 11.30 am and we arrived onsite at midday, all electronics were functioning correctly and the battery had a voltage of 13.1 V. The inverter is a EATON ELISPE ECO 500 but we swapped out the internal battery for a BSB Solar 12-200.

With a charge of 5A the battery should last at least 20 hours. But in this case it didn't even make 9 hours. It is worth noting that onsite it would have been about -5°C during the night time. I am trying to figure out why are battery is draining so fast, has anyone got any pointers ?

• I first was going to say this is off-topic, but I suppose the question is specifically about battery life as it relates to current draw, temperature, efficiency, etc. – Kellenjb Jan 6 '12 at 15:51
• Extrapolating, it looks like you should have about 80% capacity at -5C, so perhaps you want to look at efficiency or current draw. At the least you'll want to measure the load. Would it be permissible to cut off the input power while there to monitor things and monitor the battery voltage over the course of several hours? – Chris Stratton Jan 6 '12 at 16:12
• @Klonq, Are you pulling 5A on the AC side or are you measuring that current out of the battery? – Kortuk Jan 6 '12 at 17:26
• I believe the problem started at "Last night was the first time the system was put to the test..." ;) – tyblu Jan 7 '12 at 1:06
• @tyblu I thought that also, decided to keep my mouth shut though :-) – Kellenjb Jan 7 '12 at 1:07

If you were pulling directly from the battery, 20 hours at 5 amps could be reasonable. The problem is that you aren't pulling from the battery directly, instead your UPS has electronics to convert DC to AC, and then your communication system probably has electronics to convert your AC back to DC.

You can roughly assume about a 50% efficiency when operating in this manor. With that assumption, lets look a bit about the power you are using.

Lets first convert to Watts as this will be the universal unit that can be used across voltages. 5 Amps at 12v is 60 Watts. I don't know anything about your communication system, but I will assume it could be pulling 60 Watts. If you take into effect the 50% efficiency, you will be pulling 120 Watts from the battery. It is a little hard for me to see the graph, but it looks like this will almost cut your life in half to about 11ish hours.

So 11ish hours is what you should be expecting at 25*C, but you are at -5*C, to make things easy I will assume it is actually at 0*C when you account for some electronics around it keeping it a little warmer. At this temperature you should be getting 85% efficiency, which brings your life down to 9.35 hours.

This is still more then the "less than 9 hours" that you are seeing, but I have also only roughly given some values. So I would say what you are seeing real world is within tolerance of my rough calculations meaning that there is nothing wrong with your UPS.

As a side note, I usually see people place backup generators in places that they might expect power outages for extended periods of time. The battery backup is there to pickup for the brief period before the backup generator kicks on and then also handles the transition back to grid power. Also many systems implement a battery backup that accesses the battery directly as to not have the DC->AC->DC conversion. I have seen this in security systems among many other places.

• I have seen laptops plugged into a UPS on a specially made rolling cart. (at a hospital) The dumbest part is that it was strapped down so that you couldn't use the built-in keyboard. – Brad Gilbert Jan 6 '12 at 18:26

When was the last time you had the battery tested? How do you know that in its current condition that it still performs to 100% of the manufacturers specifications for a new battery.

In critical datacenters where UPS supplies are important they usually have them professionally tested on a yearly basis or perhaps more often.

If the battery is okay, I think I would expect it to last a bit longer. I'll run through the calculations:

The battery is 12V and 200Ah. Watts wise this is 2400Wh. If we say the battery is running at 80% capacity due to the low temperature we get 2400 * 0.8 = 1920Wh

If you are drawing 5A at say 50% efficiency then that's 120W. 1920 / 120 = 16 hours.

Even with considerably less efficiency or a lower temperature, under 9 hours seems a bit off.

To double check - according to the table at the bottom, 1 cell can supply ~20W for 20 hours, so with 6 cells 120W for 20 hours is expected under normal conditions (25 degrees C)

Also 5A is a low current (for this battery) so the capacity should not be affected by this. According to the capacity figures, drawing anything under 10A should allow the full 200Ah to be realised (goes down to 170Ah at 34A)

Are you sure you are only drawing 5A? How old is the battery?

• the 5A is measured by multimeter on the battery with no grid power. I am starting to think either the battery is damaged or the inverter has battery protection systems and is cutting the load at too high a voltage. – klonq Jan 9 '12 at 10:58
• The battery is about 6 months old and this is the first time it has been discharged – klonq Jan 9 '12 at 10:59