Electrical hazards of battery systems

A battery string is a DC source that is always on. HSE-321 Lesson 7 treated the battery as one of the sources to isolate before working on a UPS, and showed why stored energy in the UPS DC bus capacitors must be discharged and verified. This lesson looks at the battery from the other side: the electrical hazards of working on, in or near the battery itself, where isolation is never complete. It covers DC shock, DC arc flash and short circuits, the specific hazards of tiers, shelves and intercell connections, and the work practices and tools that control them.

DC shock

Electric shock from DC differs from AC in a few ways, but it is no less dangerous at the voltages found in UPS batteries. Direct current does not cause the same let-go problem as 60 Hz alternating current at low levels, and the threshold for ventricular fibrillation is higher for DC than for AC. At several hundred volts, however, the current through a body can easily exceed those thresholds, and DC causes deep burns at entry and exit points and can still stop the heart. A UPS battery string at 480 to 600 V DC must be treated as a lethal shock hazard, with the same approach boundaries and protective equipment as AC equipment at similar voltages.

The voltage that matters is not always the full string voltage. A worker touching two points on the same string is exposed to the voltage between those points, which depends on how many cells lie between them. Touching the two terminals of one 12 V monobloc is not a shock hazard, but touching the positive end of tier one and the negative end of tier four may expose a worker to the full string. Because racks are built in tiers with the string snaking from one shelf to the next, terminals at very different potentials can sit only a short reach apart. This is one reason battery racks have terminal covers and why work rules restrict how many tiers can be exposed at once.

Ground faults add a second path. Most UPS battery systems are ungrounded or high-resistance grounded so that a single ground fault does not trip the system. A ground fault on one part of the string can then put a dangerous voltage between another part of the string and the grounded rack frame, even though nothing seems wrong. Ground fault detection on the battery bus, regular checks and treating the rack frame as potentially energized until tested are all part of the control.

DC arc flash

When a conductive object bridges battery terminals or a fault develops in DC cabling, a DC arc can form. DC arcs differ from AC arcs in an important way: alternating current passes through zero 120 times a second, which helps an AC arc extinguish, while a DC arc has no current zero and can sustain itself once established. A battery can also supply its full fault current for as long as the fault lasts, limited only by the internal resistance of the cells and the connections and by any fuse or breaker that clears the fault.

Battery fault currents are high. A large lead-acid or lithium-ion string can deliver several thousand to tens of thousands of amperes into a bolted short circuit. The resulting arc releases heat, molten metal, pressure and light in the same way as an AC arc, and it can ignite hydrogen in a lead-acid room or damage cells and start thermal runaway in a lithium-ion system.

Arc flash incident energy for DC systems is estimated differently from AC systems. IEEE 1584 addresses AC systems only. For DC, NFPA 70E includes informative methods in its annexes, one of which is based on the maximum power transfer principle: it assumes the arc current is half the bolted fault current and calculates the energy delivered at that maximum power. This method is deliberately conservative. Other published methods model the arc more closely and usually give lower results. As with AC systems, a qualified engineer should perform the study, and the label on the battery equipment governs.

For teaching purposes, the maximum power method can be written as \(E = 0.01 \, V \, I_{arc} \, t / D^2\), where \(E\) is the incident energy in cal/cm², \(V\) is the system voltage in volts, \(I_{arc}\) is the arc current in amperes (taken as half the bolted fault current), \(t\) is the arc duration in seconds and \(D\) is the working distance in centimeters. The bolted fault current of a battery is roughly the open-circuit voltage divided by the total internal resistance of the string, including connectors.

# Teaching estimate of DC arc flash incident energy using the maximum power method.
# NOT a substitute for an engineering study. All values are illustrative.
def dc_incident_energy(v_sys, i_bolted, t_s, d_cm):
    i_arc = 0.5 * i_bolted                      # maximum power assumption
    return 0.01 * v_sys * i_arc * t_s / d_cm ** 2  # cal/cm2

v_string = 545.0                 # VRLA string at float, volts
r_cell_mohm = 0.45               # internal resistance per 2 V cell, milliohms (illustrative)
cells = 240
r_links_mohm = 20.0              # connectors and cables, milliohms (illustrative)
r_total = (cells * r_cell_mohm + r_links_mohm) / 1000
i_bolted = v_string / r_total
print(f"Bolted fault current about {i_bolted:,.0f} A")

for t in (0.05, 0.5, 2.0):       # fuse or breaker clearing time, seconds
    e = dc_incident_energy(v_string, i_bolted, t, 45.7)
    print(f"Clearing time {t:4.2f} s at 45.7 cm: about {e:5.1f} cal/cm2")

Bolted fault current about 4,258 A
Clearing time 0.05 s at 45.7 cm: about   0.3 cal/cm2
Clearing time 0.50 s at 45.7 cm: about   2.8 cal/cm2
Clearing time 2.00 s at 45.7 cm: about  11.1 cal/cm2

The results show how strongly clearing time drives the hazard. In this example a fast-acting fuse at the battery terminals keeps the estimate below 1 cal/cm², while a fault that lasts two seconds exceeds 10 cal/cm², and a fault that is not cleared at all, because it lies between the cells and the first protective device, can last much longer. A fault on the battery side of the disconnect has no upstream protection except the battery's own internal resistance, which is why work on battery terminals and intercell connections deserves special care.

Where protection is missing

In an AC system there is almost always a protective device upstream of the work. In a battery string there is not. The cells themselves are the source, and a short circuit across a few cells in the middle of a string is protected by nothing at all until the conductor melts or the arc self-extinguishes. String fuses or breakers protect the cables between the battery and the UPS but not the cells and connections inside the string. Some battery designs include fuses between tiers or within modules to divide the string into smaller sections; lithium-ion modules often have internal fuses and contactors controlled by the battery management system. Workers must know where the protection is and assume there is none at the point where they are working.

Hazardous tasks

Most battery work involves exposure to energized terminals. The tasks with the greatest electrical hazard are those where tools or conductors come close to terminals at different potentials.

• Torquing intercell connections, where a wrench can bridge two terminals or slip onto an adjacent cell.
• Replacing a monobloc or module within a live string, which requires breaking connections under the string voltage.
• Measuring cell voltages and impedance, where test leads and probes approach many terminals in turn.
• Removing and replacing terminal covers and tier barriers.
• Connecting and disconnecting strings during installation and removal.
• Working on DC cables and busbars between the battery and the UPS.

Before any of these tasks, the work should be planned with a risk assessment covering shock, arc flash and chemical hazards, and the battery should be separated into the smallest practical sections. Many battery systems provide inter-tier disconnects or link bars that can be removed to break the string into sections of a lower voltage. Opening the string at the right points can reduce both the shock hazard and the available fault current for the work, and is an example of the hierarchy of controls applied to an energy source that cannot be eliminated.

Tools and personal protective equipment

Insulated tools rated for the voltage are essential for battery work. Wrenches and socket sets with insulated handles and only the working end exposed prevent a tool from bridging terminals. Torque wrenches should be insulated or used with insulated extensions. Test leads should be rated for the voltage and the measurement category, with shrouded tips. Metal objects that are not part of the task, such as watches, rings, bracelets, necklaces, metal pens and keys, must be removed: a ring bridging two battery terminals can heat to a glowing state in seconds and cause a severe burn.

Personal protective equipment for battery work combines electrical and chemical protection. Arc-rated clothing and face protection are selected from the label or the risk assessment. Rubber insulating gloves of the right class protect against shock; leather protectors over them protect against cuts and arc exposure, though some battery tasks call for chemical-resistant gloves instead or as well. For flooded batteries and any task where electrolyte contact is possible, splash goggles or a face shield, a chemical-resistant apron and chemical-resistant gloves are required. HSE-321 Lesson 10 covered arc-rated PPE and glove classes; HSE-240 covers respiratory protection, which is relevant during off-gas events rather than routine work.

Signs, labels and boundaries

Battery rooms and cabinets must carry signs that identify the hazards: the electrical hazard, with the nominal voltage and, where required, the arc flash label information; the chemical hazard of the electrolyte; for lead-acid rooms, the presence of hydrogen and the prohibition of open flames and sparks; and for lithium-ion, the emergency response information agreed with the fire service. The electrical code requires energy storage systems to be marked with the maximum voltage and available fault current, and the arc flash label tells workers the incident energy or PPE category and the boundaries. Approach boundaries for battery terminals are set by the same tables used for AC equipment at the equivalent DC voltage, as HSE-321 Lesson 5 explained.

DC systems at lower voltage

Not every battery hazard involves hundreds of volts. Telecom DC plants at 48 V and rack-level battery shelves at around 50 V have low shock hazard, but they can deliver thousands of amperes into a short circuit. Dropping a wrench across a 48 V busbar can produce a violent arc, molten metal and a serious burn. The same rules apply: insulated tools, no jewelry, eye and face protection, and covers kept in place.