The Oft Misinterpreted Requirements of Stationary Battery Systems
When confronted with a situation about which he or she has little knowledge, an inspector or authority having jurisdiction (AHJ) will tend to take a literal interpretation of the applicable code and err on the side of safety. On the surface, this approach seems admirable. But in practice, it sometimes leads to some very strange, as well as expensive and unnecessary, corrective actions, especially in the case of battery installations for network critical physical infrastructure (NCPI). The various codes that govern NCPI installations typically only include a few paragraphs that cover batteries, and it isn't unusual for some AHJs to go years without ever having to approve a battery installation.
Contradictory information. It gets more complicated when an AHJ seeks education about battery systems, only to be confronted with seemingly contradictory information. What is true for one type of battery may not be true for another. The result is a misapplication of a governing code by an AHJ trying to do the right thing but for the wrong reasons.
For example, the National Electric Code (NEC) broadly defines a storage battery as any rechargeable electrochemical type, including nickel-cadmium (Ni-cad), lead-acid, or “other.” By contrast, the International Fire Code (IFC) recognizes one single battery technology. It has two chapters on lead-acid batteries — one for valve regulated lead-acid (VRLA) batteries and one for vented (flooded) lead-acid batteries.
When confronted with a battery site that is not lead-acid, an AHJ trying to enforce the requirements of the IFC has no guidance. For example, one of the largest stationary battery installations in North America is a Ni-cad battery system. Potassium hydroxide, the alkaline electrolyte in a Ni-cad battery, can be just as damaging as the sulfuric acid-based electrolyte in a lead-acid battery, and both types can release hydrogen into a room.
A proposal was made to amend the IFC to cover stationary storage batteries instead of lead-acid batteries. However, the governing IFC panel members rejected this proposal because they couldn't correlate the proposed amendment's use in two chapters of the IFC code. In the next code cycle, the IFC may expand its coverage to include other battery technologies. Until then, however, an AHJ will have to decide how to fit the current requirements of the IFC to the specific installation site and its characteristics.
In another example, a mechanical inspector in a large Southwestern city requires all uninterruptible power supply (UPS) systems to tie into a dry contact that will automatically shut down battery charging if a room ventilation fan fails. This requirement is meant to prevent the accumulation of hazardous levels of hydrogen. The requirement is partly based on an instance of an explosion in an unoccupied vented-lead-acid battery room where all safety and ventilation equipment had been disabled. This interpretation isn't popular with information technology (IT) and facilities managers who may have spent hundreds of thousands of dollars creating highly reliable, redundant power systems, only to put their mission-critical loads at the mercy of a failed fan or relay. This mechanical inspector required that this feature be included even when given data proved it would take weeks or even months for gas to accumulate to a hazardous level within the volume of the room — certainly enough time to alarm and correct a fan failure. Today's UPS systems can detect and prevent conditions that would allow hydrogen release in the first place, but the contact closure with charger disconnect is still the rule in that city.
This same AHJ makes no distinction between VRLA batteries, which are sealed and release little or no hydrogen under conditions of normal use, and vented batteries, which continually release up to 60 times more hydrogen gas than a VRLA battery under float-charging.
Finally, this misapplication of the governing code by the AHJ is applied to any room that contains batteries, including a data center. The fire code doesn't put restrictions on battery systems with less than 50 gallons of electrolyte, but the mechanical code is mute on the point, so the mechanical inspector puts this rule into effect even on VRLA battery systems with as little as five or 10 gallons of electrolyte. Although his/her decision seems to meet the letter of the code, it misses the intent entirely.
In still another example, the AHJs in many cities require data centers that contain batteries to duct exhaust air to the outside of the building, even though the closest place to do so might be 10 floors away and the room holds sealed, non-gassing batteries. Also, IT environments have specialized computer room air conditioners that have much higher levels of humidity control and dust filtering than ordinary comfort air systems. They also recycle as much air as possible to avoid contamination with outside air. Requirements for special ducting seem to defeat this objective.
An inspector's interpretation can be justified by the NEC Handbook, which seems to say that one should design a facility for worst case failure conditions, in seeming contradiction to the Uniform Fire Code, NFPA-1, which states in Article 4.1.3.3.2.1 that “storage, use, or handling of hazardous materials in a building or facility shall be accomplished in a manner that provides a reasonable level of safety…during normal storage, use, or handling operations and conditions.” The IFC, in Section 609, permits VRLA batteries to be installed “in the same room with the equipment they support,” provided the system includes “a listed device or other approved method to preclude, detect and control thermal runaway.” Mechanical codes make no such accommodation.
Despite changes to the latest versions of both the Uniform Fire Code and the IFC, some AHJs continue to require users to install spill containment barriers around non-spillable VRLA battery systems. For example, at a central telephone office, this requirement cost one customer more than $100,000 to uninstall and re-install equipment. That doesn't include lost revenue during the several days that the central office wasn't in service. In another case, the containment barrier obstructed the opening of a battery cabinet door. In still another case, it made removal of a depleted battery nearly impossible. Finally, at a West Coast site, the AHJ decided that the UL Listing on a piece of IT equipment that contained batteries didn't apply if the room in which it was installed failed to meet all of the requirements of NEC Art. 645 Information Technology Equipment Rooms. The supplier was then required to re-wire interconnect cables inside the listed equipment.
Well-meaning but troublesome. The message of the above anecdotes isn't that AHJs are uninformed or mean-spirited. In every instance, the AHJ was doing what he or she believed to be the safe and correct thing to do. And in every instance, the decision was based on misinformation or misunderstanding of the particular technology.
Even though batteries have been widely used for more than a hundred years, they're still misunderstood and feared by many. As new battery technologies, such as stationary lithium batteries, become commonplace, AHJs will continue to look to the various governing codes for guidance. But if these codes don't address these new technologies, the AHJs will continue to do as they have in the past.
So how does a plant/facilities manager, IT manager, or installing contractor match their installed stationary battery systems with perceived code requirements? The best approach is to have the AHJ fully explain his or her objection and the basis of the objection. Armed with this information, you should contact specialists within the manufacturing community who deal with code issues and the specifying engineer. The response can — and should — be addressed on the basis of case studies and instances in which similar concerns were expressed and were satisfactorily answered within the various code constraints.
McCluer is external code compliance manager with American Power Conversion in New Kingston, R.I.
