[RE-wrenches] Battery venting issue

Thu Feb 6 07:30:19 PST 2014

Allan,

The primary concern should be the volume of gas discharge if over voltage occurs. This study helped me understand the risks: http://www.sonnenschein.org/Gassing.htm

Quote from the study: "Gassing according to IEC 896/2-Oct 95 @ 2.3v / cell within 30 days was measured @ 5mL/cell/ah while Overcharging @ 2.48v/cell caused 900m L / cell / ah within 30 days!"

Also heres a technical bulletin from C&D that discusses how long a 2% concentration takes during overcharge. See document 41-6739: http://www.cdtechno.com/resource/support_doc.html

Thank you,

Larry Crutcher
(928) 342-9103
www.starlightsolar.com
larry at starlightsolar.com

Retail Store & Warehouse
2998 Shari Ave.
Yuma, Az 85365

USPS Postal Mail Only
11881 S Fortuna Road, #210
Yuma, AZ 85367

On Feb 5, 2014, at 7:38 PM, Allan Sindelar <allan at positiveenergysolar.com> wrote:

Wrenches,
I need a bit of help here if you have it. Since 2002 we have installed somewhere between 30 and 35 systems with sealed batteries installed in manufactured enclosures, originally Outback enclosures and in recent years Midnite MNBE enclosures. At least ten of these have been indoors in one form or another - usually a laundry or mechanical room. Our battery of choice is Concorde SunXtender. We have only added mechanical ventilation (Zephyr Power-Vent to outside) if the battery enclosure itself is sealed. Nearly all of these have been permitted and inspected systems, and we have never     had a problem with the inspectors. Of course, we always vent flooded systems to the outside, nearly always using a Power Vent fan.

Now we have. An AHJ failed a system for lack of ventilation, and our attempts to resolve it have not been effective. The Chief Electrical Inspector has weighed in, and we are right at the point of filing a Request for Code Interpretation with the New Mexico Electrical Division Technical Advisory Panel. 

I have not wanted to just add ventilation to pass inspection because of the precedent doing so is likely to set for future installations. The GC on the job supports my attempts to push back, as do the homeowners. The Chief Inspector thinks that the 700 square foot unheated room in which our system is installed is a bedroom; it's actually a storeroom for the homeowners' collectible book home business.

My request: please send me documented work by others establishing that PV systems with sealed VRLA batteries are used specifically because they are considered safe without venting to the outside. If you know of good online links, I could use them too. For example, the AHJ asked for a document stating that the batteries or the enclosure were specifically approved for this use in an indoor location. I can't - Midnite battery enclosures are simply listed to UL508A, which is "industrial control panels" and there's nothing specific to this application in the standard.

To me this is a common-sense issue, but common sense doesn't cut it when needing to prove a procedure. Can anyone help?

For what it's worth, or for those Wrenches with too much spare time, below is the text of the original defense of our installation that I sent to the AHJ. His response was that he's not an electrical engineer and this would have to be taken upstairs. For what it's worth, I'm not an EE either... My frustration is showing, I'm sure.

Thank you for any links, reports or other resources you may be able to send my way.
Allan

-------- Original Message --------

Mr. [AHJ],
I have done some research as followup to our discussion last week about battery venting for the [X] job. Here are several perspectives on the issue:

The NEC Section 480.9(A) states only that "Provisions shall be made for sufficient diffusion and ventilation of the gases from the battery to prevent the accumulation of an explosive mixture". At root, you are questioning whether ventilation of the batteries into the storeroom at the [X] home is sufficient under worst-case conditions.

The NEC Handbook entries for Section 480.9(A), which are considered as explanatory support documentation and are not Code requirements, include two paragraphs that are fundamentally contradictory to each other. The two read: 
The intent of 480.9(A) is not to mandate mechanical ventilation. Hydrogen disperses rapidly and requires little air movement to prevent accumulation. Unrestricted natural air movement in the vicinity of the battery, together with normal air changes for occupied spaces or heat removal, normally is sufficient. If the space is confined, mechanical ventilation may be required in the vicinity of the battery.
This paragraph refers to batteries in general, including flooded batteries which release hydrogen gas as a normal part of the charging process. The Handbook section goes on to specifically identify sealed batteries as being unlikely to release explosive gases:
Although valve-regulated batteries are often referred to as "sealed," they actually emit very small quantities of hydrogen gas under normal operation and are capable of liberating large quantities of explosive gases if overcharged. These batteries therefore require the same amount of ventilation as their vented counterparts."
Well, no, not exactly. Valve-regulated batteries may indeed require the same amount of ventilation, but not for the same purpose or under the same conditions. 

Flooded batteries release hydrogen gas as a normal part of every charge cycle. While it is unlikely that the hydrogen gas could accumulate to the 4% concentration to become combustible, given its natural dispersion, the hydrogen sulfide released with the hydrogen gas is an unpleasant irritant and is potentially toxic with prolonged exposure at high concentrations. Because of the normal gassing during the charge cycle, we always provide ventilation of these gases to the outside. With sealed batteries, the purpose and intent of ventilation is not to ensure ventilation during the normal charge cycle, but rather to ensure the safety of the dwelling and its occupants in the event of a catastrophic failure resulting in the "worst-case scenario" of unregulated overcharge. In actual experience, the charge regulator (from the PV array) and the inverter/charger (from a backup generator in an off grid home) are the bottlenecks through which all charge current must pass, and failures invariably occur in an "open circuit" mode, rather than in a "closed circuit without charge regulation" mode.

Nevertheless, we must accommodate the most hazardous potential outcome, which would be unregulated overcharge of an already full battery during periods of high insolation (or the equivalent input from an engine generator). In order to determine the expected amount of hydrogen gassing under worst-case conditions, I contacted my Concorde distibutor, Marc Kurth of Centex Batteries, LLC in Bastrop, TX, 512 308-9002. He in turn spoke with the engineering department at Concorde Battery, the manufacturer of the batteries used in the [X] PV installation. Their analysis of calculated gassing and airflow rates is in the attached pdf document which they provided to us. The batteries in the [X] PV system are Concorde SunXtender PVX-9150T, rated 915 amp-hours at the C/24 rate. There are 12 cells in a single series string of 24 Vnom.

The storeroom in which the PV system is located has interior dimensions of 19' by 37' by an average of 10' tall, or approximately 7,000 cubic feet. It's a large open space. The room has four Pella double-hung windows, each rated by the manufacturer at 0.3 cfm fenestration, or 1.2 cfm for all four. Each exterior door (the third door to the interior living space is excluded as a conservative calculation but also adds to overall ventilation) is rated at 0.6 cfm, for a total of 1.2 cfm for the two doors and 2.4 cfm for the building, assuming no other openings of any sort, such as for wires or for natural convective losses due to any other air leakage or roof ventilation. 

The 2,000 watt PV array will provide at most about 65 peak amperes of DC current into the batteries, for the equivalent of a cumulative daily total of around seven hours in summer. (The inverter/charger is capable of feeding 105 amperes into the batteries from a generator, but by the specific stated preference of the homeowners, the home does not have a backup generator and does not include the ability to accept generator AC input.) Assuming the worst case of 75 amperes flowing unregulated into this 900 ampere-hour battery, this C/12 charge rate is capable of raising the batteries to 30 V DC, or 2.50 volts per cell (vpc). The cell voltage will not rise about this level because the internal resistance of the battery, which increases as the voltage increases, prevents it. Note also that 75 amperes is a peak current that could only be maintained at midday during conditions of cold, dry air when the solar insolation intensity is well above standard test conditions (STC) of 1,000 watts/square meter, when the sun is perpendicular to the array. As the sun passes across the sky, the available output current drops substantially. At a reduced input current, the maximum vpc drops to around 2.40 vpc (and continues to drop thereafter) and the maximum temperature also drops, in which case gassing reduces by a factor of about 20 below the rate at 2.50 vpc.

As an additional factor in our calculations, note that all modern charge controllers are designed to receive PV input at a higher voltage and lower current than the nominal battery voltage, converting this to higher current at the lower actual battery voltage. The Midnite Classic charge controller in this application works this way. In a closed-circuit failure of the charge controller's functions, the higher array voltage and lower current would pass through to the batteries. As long as the input voltage is higher than the battery voltage, the batteries will accept current, but additional voltage does not increase the current into the batteries or the amount of hydrogen released. Rather, in this case the PV modules, which are wired as four strings of two modules each, will not exceed the rated short-circuit of the modules x 1.25 (per NEC for PV source circuits. With four strings, this is (8.61 x 4 x 1.25 =) 43.05 amperes. This is less than half of the maximum input current used to calculate worst-case input (as shown in the following paragraphs), and as such is unlikely to be sufficient to raise the cell voltage to even the level calculated.

Per the attached engineering analysis by Concorde, assuming that at a sustained 2.50 vpc the temperature of the batteries rises to 50ºC (122ºF), the amount of hydrogen released at a constant current at 30V DC, or 2.50 vpc, at 50ºC is 5.6 cc/hour/Ah/cell. This converts to (5.6 x 915 x 12 =) 61,488 cc of hydrogen released per hour. Converting cubic centimeters to the more useful cubic feet, 61,488/21,317 cc/cuft = 2.17 cubic feet per hour of gas released. This amount is less than the total fenestration of that room (not including the door to the living space) of 2.4 cubic feet per minute, or (2.4 x 60 =) 144 cubic feet per hour of natural leakage to the outside through closed windows and doors. 

To take this one step further, 2.17 cubic feet is 0.031% of the volume of the storeroom. It would take 30 times this concentration to exceed 1% by volume in an airtight container. 4.1% concentration is the threshold at which hydrogen gas becomes combustible.

Also at 2.50 vpc, at 50ºC, the airflow required to keep hydrogen accumulation below 1% is 0.0093 liter/minute/Ah/cell, or [(0.0093 x 915 x 12)/28.32 liters/cubic foot =] 3.6 cfm, or 216 cubic feet/hour. While this exceeds the default window and door fenestration of 144 cubic feet per hour, it is sufficient to disperse hydrogen. Note that these batteries are not in a confined space; the batteries are located in a space of 7,000 cubic feet. Note also that is the threshold for staying below 1% hydrogen concentration; 4.1% is the threshold at which hydrogen becomes explosive.

I reviewed our records pertaining to the use of sealed batteries in residential off grid PV systems and in grid-tied PV systems with battery backup. We have installed more than thirty such systems, although the great majority have been installed since 2005. Of those, I have identified at least nine permitted and inspected systems in which the batteries have been located in what may be considered enclosed spaces without ventilation between the interior space and the outside air. Indeed, several of these are in spaces much smaller that the Shutt storeroom. This is the first time in which an AHJ has expressed concern about adequate ventilation of sealed batteries.

In two of these thirty-plus confined interior installations, the sealed batteries were installed in custom-made sealed enclosures which were wrapped in sheet plywood with controlled intake ventilation. In both of these we purposely installed Power Vent battery fans (as we install in all of our systems with flooded lead-acid batteries) ducted to the outside as a safety feature to prevent the possibility of accumulation of gases within the battery enclosure itself. However, in all of the remaining systems we have used manufactured steel battery enclosures Listed to UL508A. Ventilation from the cabinet into the room where it can dissipate has always been considered to be adequate in these applications.

I believe that I have conclusively established that in a worst-case scenario, the batteries cannot release enough hydrogen to come even close to dangerous levels. In practical terms, if a failure were to occur when the residents were away, the batteries would be permanently damaged by a failed controller, but no danger exists to the home. If the residents are present when the failure occurs, they would in short order smell the "rotten egg" smell of hydrogen sulfide. Following their noses, they'd find a much stronger smell in the storeroom, suspect that the batteries were the source, turn off the circuit breakers on the system (which are readily accessible per NEC) and open the doors or windows. 

The 2011 NEC Hanbook states, as noted above: "If the space is confined, mechanical ventilation may be required in the vicinity of the battery." The storeroom at the [X] residence is simply not a "confined space" as built.

Thank you for your consideration of this defense of our installation practices.
Allan Sindelar
-- 
Allan Sindelar
Allan at positiveenergysolar.com
NABCEP Certified PV Installation Professional
NABCEP Certified Technical Sales Professional
New Mexico EE98J Journeyman Electrician
Founder, Positive Energy, Inc.
A Certified B CorporationTM
3209 Richards Lane
Santa Fe, New Mexico 87507
505 424-1112 office 780-2738 cell
www.positiveenergysolar.com

<Gassing Rates.Technical Note.Rev.10-29-08.pdf>_______________________________________________
List sponsored by Home Power magazine

List Address: RE-wrenches at lists.re-wrenches.org

Change email address & settings:
http://lists.re-wrenches.org/options.cgi/re-wrenches-re-wrenches.org

List-Archive: http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org

List rules & etiquette:
www.re-wrenches.org/etiquette.htm

Check out participant bios:
www.members.re-wrenches.org