[RE-wrenches] Fwd: single string of batts: must the interconnects beinverter size cables?

[Michael Welch]

From: jfh at greenhousepc.com


Solar Ray writes:

>> These are grid tie with battery backup systems. All inspected.
>>
>> Personally, I think that we should have OCP on each series string, but that's a whole nother can o' worms.
>> Most sealed batteries (and golf cart batteries) are fairly current limited. You need at least 2 parallel strings
>> to even get close to meeting the inverters surge requirements. (especially in 24 v systems)
>
>This is true for 24 volt systems with small banks of small batteries. 
>It is absolutely not true for 48
>volt systems where Vdrop is 1/4th what it is for a 24 volt system, and
>1/16th what it is for a 12 volt system.
>T-105s and similar sizes are pretty small, but there are sealed
>batteries with the capacity and plate
>surface area to deliver substantially more current.
>
>It's also a matter of where the battery is operating at the time you try
>testing Psurge.  All batteries have a
>problem with "coup de fouet" / "the crack of the whip" (the sharp drop
>after an abrupt increase in load from 
>a low or no-load condition), but most will recover significantly as the
>battery adjusts to the load.  Based on
>that, the worst-case scenario is a steadily increasing load as the
>voltage will fall much more slowly than
>with a single large load applied at once.  I've achieved C/1 in
>discharge on AGMs during tests, and if I'd had
>larger shunts, I would have tried higher rates of discharge.  This was
>for far longer than the 5 seconds you
>mentioned elsewhere.
>
>I have a test coming up for a client where I'm going to try and get a
>pair of GVFX3648's to sell 7.2KW.
>I'll try to remember to capture some voltage plots and post them
>somewhere.  The reduction in voltage drop
>is significant as the load is increased, compared to all at once.  I had
>those inverters at 2/3rds capacity
>and still above 46 volts.  I was able, just now, to run both at 100%
>(and then the chargers turned on due to
>low voltage, because I'd not turned them off ...) and keep the voltage
>above 46.4V -- here's the plot:
>
>http://www.greenhousepc.com/GVFX-Sell-All.gif
>
>Here's the numerical meter info --
>
>STATUS/FX/METERS-P01
>inverter     0.0 aac
>current     31.0 max
>DOWN  UP   TOP  PORT
>
>STATUS/FX/METERS-P02
>inverter     0.0 aac
>current     30.0 max
>DOWN  UP   TOP  PORT
>
>STATUS/DC/BATT---P04
>battery min today
>          46.6 vdc
>DOWN  UP  RESET  TOP
>
>This is compared to clients' systems were they've had more transient
>loads (in the range of 4.5KW for a
>pair of VFX3648) and reached lower voltages (in the 44's and 45's) with
>larger battery banks.
>
>> Main cables and breaker are sized as always, (4/0 with a 250 A breaker) but parallel string jumpers are
>> reduced to 2/0 as they are carrying 1/2 to 1/3 of main current.
>
>Having seen all but one string of batteries fail in a multi-string
>system, this is not a safe assumption.
>You cannot assume all of the strings are working properly, or even that
>only one string fails.
>--
>Julie Haugh
>Senior Design Engineer
>greenHouse Computers, LLC // jfh at greenhousepc.com // greenHousePC on
>Skype