Paralleling one 'A' battery to another 'A' battery 3

[marks1080]

Hi folks, looking for some insight from everyone...

I'm replacing all of the DC station service equipment at a large 230/115/27.6 station. I'm pondering different outage scenarios related to different design options available to me. Let's assume that the old DC stn srvc (the one to be replaced) is fully in service, and my new DC stn srvc is installed (ATS stacks, batteries, Dist. panels, etc...) but not feeding any load. Does anyone see a problem for briefly paralleling the old 250VDC 'A' battery with the new 250VDC 'A' battery for brief periods of time? This would really help reduce outage times. I know you would never want to parallel and 'A' battery to a 'B' battery, but I've never been in a situation where I technically had two separate 'A' or 'B' batteries available.

Both the old and new 'A' batteries are 250VDC floating.

Personally, I can't see any reason why I wouldn't be able to do this, but when I talk about it, it makes some of the 'higher-ups' nervous. The specific benefit for my job would be drastically reduced outages at this station, which is quite large and feeds the better part of a large city.

Any insight would be appreciated. Thanks!

 

[davidbeach]

If it's the 'A' battery, that implies the existence of a 'B' battery. To me that would imply a certain level of redundancy; if fully redundant you should be able to have one out at a time.

We often connect a temp battery in parallel with station battery so the station battery can be disconnected for maintenance. How are you presently doing battery maintenance?

 

[marks1080]

It's an old station and even though there are separate 'A' and 'B' battery systems not all systems at all voltage levels are fully redundant. I don't know what the battery maintenance practice is for this particular site, but i will be sure to inquire. This job is also just one part of an overall station rebuild, and due to circumstances outside of my control I will have to pick up old equipment on the new system which should be replaced prior to my work. So I'm trying to accommodate the old loads, as well as what i know the future loads and distributions (based on current standards)should be.

When you connect the temp battery to your existing battery is it a make before break connection? I assume it is. If so, are there any restriction on how long you can keep the systems parallel or is that 'don't care?'

Considering that two 'A' batteries may not be exactly the same 250VDC, if there is a slight difference between the two would it cause any issues for the chargers connected to each system or will the paralleled system voltage normalize between the two batteries?

 

[davidbeach]

So I'm just the customer for the DC station service, 125V in most cases, not the guy that has to make sure it's always there. Maintenance is always make before break, no DC outages.

I'd also expect/hope that your project is changing all the DC panelboards, probably adding a significant number of additional circuits (that's what our older systems would need); why not just move equipment to the new DC one device/position/rack at a time - a multitude of small outages rather than one station wide outage - if you can't the whole system out a once without a station outage?

 

[ScottyUK]

If the batteries are of the same technology (Plante? Flooded NiCad?) and the same cell count then it shouldn't be a problem if both batteries are fully charged. Multiple parallel strings are frequently seen in UPS installations. Goes without saying that a 250V battery should be treated with caution - check and check again before throwing the paralleling switch.

 

[submonkey]

Hi,

If you have access to the battery cell
datasheets, they will tell you the internal
resistance of the cells.

From the internal resistance and a measured
voltage difference, you can calculate the maximum current which will flow when the
batteries are paralleled. If you can keep
this current to a reasonable value (compare
with charging/load numbers), you can parallel
the batteries. If you don't have a datasheet
for the old battery, you can probably measure
the resistance with a dummy load and voltmeter.

Be very, very careful when paralleling. Measure the difference across the poles
of the paralleling switch before you close
it. Use appropriate fusing. A large sub
battery can produce significant arc energy.

Thanks,
Alan

 

[FreddyNurk]

Submonkey is right, whilst quite a number of battery installations consist of multiple parallel strings, what you don't want to do is exceed the breaking capacity of any distribution system if you end up with a fault. Some interesting advances have been made in the last few years in terms of protection of DC systems (such as the issues surrounding polarised DC breakers), thus creating one giant battery bank is not necessarily a good thing, although it certainly can be done within limits.

Connecting parallel strings of batteries in different states of charge can get awkward. A staged transition for the connected equipment, with due consideration of planning, and inadvertent connection of systems together would probably be the best option. If the existing battery system has the option for separating the battery strings, then staging equipment that way may allow for paralleling of the different banks without causing an issue in terms of potential fault current.

Don't forget that DC is much harder to break as it doesn't have a zero crossing, and batteries typically don't have a shutdown button either.

 

[DanEE]

"If the batteries are of the same technology (Plante? Flooded NiCad?) and the same cell count then it shouldn't be a problem if both batteries are fully charged. Multiple parallel strings are frequently seen in UPS installations. Goes without saying that a 250V battery should be treated with caution - check and check again before throwing the paralleing switch. wink"

Ditto.. One technique we always used in equivalent paralleling situations was to place a resistive load (to limit current) across the paralleling point as a final check, to see what level of current, if any, wanted to flow through the load, before paralleling.

One tool I made that had ideal electrical characteristics for the above resistive load testing was to construct the load using quartz halogen lamps as the electrical test load. The non linear resistance with respect to filament temperature of an incandescent lamp makes an ideal test resistance for this situation. When the filaments are cold, resistance is very low and increases as the lamp begins to glow.. For 48 volt DC plants, I used 120VAC and 12VDC halogen lamps respectively. Measured cold resistance of a single 120vac halogen lamp (I forget the wattage rating of the lamps used) is around 14.5 ohms. As the final check with lower resistance, a 12VDC halogen lamp that has a cold filament resistance of 0.9 ohm was placed across the point in question. If the final check with the 12VDC lamp along with a current probe, was good, then it would be OK to proceed with paralleling. And of course if either lamp glowed that was warning enough to recheck everything.

 

[marks1080]

Thanks everyone! Some great info here. Much appreciated. I'll try to get datasheets for the old and new cells. They are both flooded NiCad type.

I'm still not sure if this will be an option for me for a variety of other reasons, but it does seem like it could be a potential option that I can present to the asset managers.

Thanks again,
Mark

 

[Compositepro]

I do not believe that the internal cell resistance is very relevant to paralleling batteries, unless a battery is defective, which will show in a voltage measurement. Batteries are electrochemical devices and there is a voltage threshold between charging and discharging. If you have two identical batteries, which should always be the case when paralleling, a full battery cannot charge a partially charged one because the charging voltage required is higher than the discharge voltage of the other battery.

 

[submonkey]

Hi Compositepro,

Your post is very interesting, and makes
perfect logical sense for batteries with
the same chemistry. It is worth noting that
cases arise which require short-term
paralling of dissimilar batteries
(changeover, etc).

Do you know a way to quantify the current
which will flow when batteries are paralleled? I don't recall seeing any V vs I curves in
data sheets, but perhaps I haven't looked
closely enough. The internal resistance
calculation I thonk should still give a
maximum - the real current will be less
because of the effect you describe.

Thanks,
Alan

 

[DanEE]

Quote: Do you know a way to quantify the current which will flow when batteries are paralleled?

Better yet, a technique to minimize the current... Adjust the chargers still on one set of batteries to exactly match the batteries to be paralleled in and let them acclimate for a while for the reason in quoted in the note below. Same voltage on each, no significant current flow.

One caveat: On old, aging charger systems using mechanical pot controls to set voltage, we didn't touch these if at all possible.. Have had them fail on movement of the pot with charger either shutting down or going wide open, best to do adjustment on the new charging system, if being installed, to match old.

This was the method used for these situations a) complete new DC power plant, including new DC distribution panel, and b) situation where new chargers, batteries being installed and had to use existing power board and c) new batteries being paralleled in either permanently or during hot switch over. All scenarios requiring no interruption to system loads.

Situation A required hot tapping existing loads over one by one to new distribution panel, new loads of course powered by new panel. Situation B required both charger systems systems being online together during hand over of load from old to new chargers. Situation C) business as usual.

I agree in part with CompositePro's comment regarding internal cell resistance not being the significant issue when paralleling. This view is supported by Alber White Papers

http://www.alber.com/

(and personally observed in measurements on close to 100 different systems) is that internal cell resistance does not change that much over the acceptable state of charge range.

(I almost hate to make the above statement for fear of diverging into the decades long battery test methodology debate of (AC) impednance, (DC) cell resistance, acutal load test results

http://www.battcon.com/PapersFinal2003/AlberPaperFINAL2003.pdf

). I can only say this from experience with Albers measurement methodology and same comment in their white papers.

As CompositePro states one battery system (not connected to charging system voltage) will not significantly change the chemical state of charge BUT, below is the issue..

quote from Alber and same statement made in C & D Technologies and other battery manufacturer white papers..

Note: quote 1 from Alber : The cell’s electrical makeup also includes a huge capacitor, which is shown in the simplified electrical model in Figure 2. The typical value of the capacitor is 1.5 farads per 100 ampere-hour of cell capacity.

quote 2 from C & D Technololgies "Capacitor value is substantial 1.3 to 1.7 Farad per 100 ampere hours of battery rating"

The stored equivalent capacitor energy in a good battery system can be enormous and when connecting two battery systems is best for those two systems to be floating at the same voltage.

 

[marks1080]

DanEE: Great info. Very relevant to my situation. Thanks so much!

 

[Compositepro]

In my experience with lead acid batteries there has never been any concern in paralleling two bateries regardless of state of charge. Dead batteries have high resistance and as they charge their voltage rises rapidly, so the voltage difference with another battery becomes small. Paralleling batteries is commonly done without incident.

 

[itsmoked]

Right up until one cell of one battery shorts..

Keith Cress
kcress -

http://www.flaminsystems.com

 

[davidbeach]

My experience with Flooded Lead-Acid batteries is that an open cell is far more likely than a shorted cell. Other types may have other predominate failure modes but no other type is appropriate for the application.

 

[ScottyUK]

Flooded NiCad is technically fine for the application, but economics weighs in favour of flooded lead-acid at least in terms of initial cost.

 

[itsmoked]

Interesting david. My experience is the exact opposite. I've never seen a open failure only shorts.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[davidbeach]

Are you saying that flooded NiCads don't have the memory problem that sealed NiCads do?

 

[DanEE]

Personally, I can't see any reason why I wouldn't be able to do this, but when I talk about it, it makes some of the 'higher-ups' nervous. The specific benefit for my job would be drastically reduced outages at this station, which is quite large and feeds the better part of a large city.

One last comment I wanted to make regarding the above ... On a job of this magnitude, the words we used to management was always "you need to prepare on the presumption the system will go down" as there is always the possibility of the unforeseen problem. Most importantly, it gives them the opportunity to put in place any necessary contingency plans, chose least disruptive time frame to do the work, and most importantly makes them inescapably responsible for the decision to proceed on the plan, given any potential risks. That includes making sure they know this is a job that should be done by experienced DC plant people who have done this type of hot work before and not by someone neither trained or with experience in this type of work, trying to save a few dollars.