The Joint Commission clarifies its generator testing requirements


September 1, 2007

Tests may be combined in certain situations

As of July 1, to be in compliance with EC 7.40, hospitals needed to have completed a four-hour generator test at some point within the past 36 months—and perform a similar generator run every 36 months thereafter.

This testing approach follows the National Fire Protection Association’s (NFPA) standard NFPA 110, Emergency and Standby Power Systems.

Generator testing is a big deal these days: In interviews with Briefings on Hospital Safety, recently surveyed hospital safety managers nationwide say that Joint Commission representatives requested to see generator testing records. That’s not a surprise to David Stymiest, PE, CHFM, FASHE, a senior consultant for Smith Seckman Reid, Inc., based in Nashville.

Stymiest says he has heard that surveyors have zero leeway on that count, as they’ve been instructed to check generator testing wherever they go.

“So a surveyed facility can be pretty sure [it] will have to produce generator test records,” says Stymiest, who is chairperson of NFPA 110’s Technical Committee on Emergency Power Supplies.

Most hospitals that Stymiest talks to satisfy the test requirement via a load bank (see “Tech talk: Details about load banks and 30% nameplate” below for more). However, some hospitals have used what is called a “dynamic load”—switching the emergency loads over to a generator and performing a test to verify that the system does indeed power its emergency loads.

What tests meet which requirements?

Besides the four-hour test, EC.7.40 also requires the following:

  • Generator tests 12 times per year for 30 minutes that draw at least 30% nameplate, or energy capacity, of the equipment (if a generator doesn’t hit 30%, the hospital may substitute a supplemental test at various nameplate ratings totaling two hours, as outlined in EC.7.40)
  • Automatic transfer switch tests 12 times per year

    The question that hospitals had after going through the four-hour test: Does that generator run also satisfy the requirement for monthly 30-minute tests, as well as the potential annual, two-hour generator tests?

    Before 2007, the answer was no. After The Joint Commission published a clarification in its July Environment of Care News, the answer is yes, provided that for the duration of the four-hour test, it draws at least 30% nameplate.

    This option applies only if you’re running diesel engines; other types of generators have no 30% requirement.

    Actual incidents can count as tests

    Further, if an actual power outage occurs that lasts for at least four hours (and draws at least 30% nameplate for diesel generators), that counts for the required four-hour test and for whatever shorter tests that might be eligible, The Joint Commission says.

    For example, if it’s been more than 20 days since your last half-hour test, an outage lasting five hours that draws enough power would satisfy requirements for monthly, annual, and triennial tests.

    But if you’re doing a four-hour load bank that draws enough power, it satisfies only the two- and four-hour requirements, because the monthly testing requirements also include transferring the facility’s power over to generators to confirm that the transfer switches still work.

    Other events can substitute as tests, too, The Joint Commission says in its clarification. Hospitals running their emergency power supply systems for four hours as part of a peak-shaving agreement with a local utility may also satisfy the triennial test requirement, provided they meet all of the other provisions of EC.7.40.

    Hospitals ask for clarification

    If all this sounds confusing, you’re not alone, says Dan Chisholm, emergency power supply systems consultant for the Motor and Generator Institute of Winter Park, FL. Chisholm is also a member of the NFPA 110 technical committee.

    The Joint Commission initially required separate two- and four-hour tests. After hospitals expressed their discontent, The Joint Com-mission and American Society for Healthcare Engineering (ASHE) polled the NFPA technical committee about a formal interpretation of NFPA 110. Combining tests is fine with committee members, as long as the requirements for each are similar.

    “Overwhelmingly, we voted to tell [The Joint Commission] that one test would be sufficient,” Chisholm says.

    Because of the way that ASHE and The Joint Com-mission worded the question, the NFPA could not issue a formal interpretation of the standard, he explains.

    However, the technical committee’s stance was clear enough to The Joint Commission that it issued the clarification, with the caveat that future NFPA action could alter the situation.

    The NFPA technical committee is expected to streamline the next edition of the standard, scheduled for 2009 release, says Chisholm. He hopes that The Joint Commission and the Centers for Medicare & Medicaid Services will follow the new standard as written, without additional rulemaking.

    “We hope to simplify it so that it will be easy for anyone to understand and won’t be so dadgum convoluted,” Chisholm says.

    Generator records need to be in order

    Surveyors look for a few basic items in your generator testing documentation:

  • That the hospital conducted the testing
  • That testing lasted for the prescribed interval
  • That the load bank or actual loads from the facility drew enough to satisfy the 30% requirement for diesel generators
  • That the hospital operated transfer switches properly

    However, there are more actions that a hospital can take to move beyond the basics and show The Joint Commission that the facility has developed a sound emergency power management plan.

    In September, 2006, The Joint Commission issued a Sentinel Event Alert that outlined preventive steps to help avoid problems caused by power disruptions.

    The Alert discussed a number of testing measures that The Joint Commission recommends hospitals carry out, and offered examples of clinical contingency plans that hospitals could put into place during a power loss.

    Examples included:

  • Accessing drugs marooned in automated dispensing machines
  • Using two-way radios in a facility as communication backup
  • Maintaining communications with hospital leaders who are off-site

    Safety managers who read that Alert might find clues to what surveyors look for, Stymiest says. The big points are to address any problems that generator testing reveals and put detailed thought into handling realistic power outage scenarios—including those caused internally by equipment failures and externally, such as by weather -conditions.

    “It’s asking the organization to go a little bit further and look beyond just the numbers,” Stymiest says of the Alert. “Make sure the lessons learned from whatever happened during the testing are investigated and followed through . . . to improve the reliability of the overall power system.”

    In his ASHE monograph, Managing Hospital Emergency Power Systems: Testing, Operation and Maintenance, Stymiest points out that during monthly testing, emergency power systems can fail and transfer switches might break. It is better for testing to uncover those problems than to find out during an emergency, he writes. The monograph, available for free on ASHE’s Web site (, is well worth reading.

    Elements of emergency power testing programs

    The new provisions of EC.7.40 should prompt hospitals to review and reevaluate their current emergency power plans. The following are three elements important to every emergency power system management plan:

    1. Emergency power gap analysis—Figure out where you need power most critically during outages of several different severities and how to get that power to those critical areas

    2. Emergency power vulnerability analysis—Determine the weak spots of your generators

    3. Risk management plan—Take those identified vulnerabilities and come up with solutions or mitigations for them

    Source: David Stymiest, PE, CHFM, FASHE, senior consultant for Smith Seckman Reid, Inc., based in Nashville.

    Tech talk: Details about load banks and 30% nameplate

    Hospital safety managers and facility directors might wonder why The Joint Commission is keen on running generators at a minimum of 30% capacity during load bank or real-world testing.

    That mantra actually comes from generator manufacturers, who know diesel engines and the way they run, says Dan Chisholm, emergency power supply systems consultant for the Motor and Generator Institute of Winter Park, FL.

    “If you can’t run your diesel engines at least 30%, you’re going to cause engine damage,” Chisholm says.

    That’s because at lower levels of output, unburned fuel accumulates in the engine, a condition called “wet stacking.” At 30% capacity or higher, the inside of the engine gets hot enough to burn all the fuel piped through it.

    The 30% rule also relates to another concept that might confuse hospital safety managers new to the technical side of emergency power: using a load bank versus the actual hospital power load for generator testing.

    Some people might assume that a hospital would bring in a load bank—basically, a power-eating device on a trailer in the parking lot—because it’s easier than arguing with clinicians and administrators about what would be the best time to test the generators.

    That’s not true, though, Chisholm says. The reason most facilities use a load bank is that regular power loads at a hospital won’t necessarily tax the emergency system enough to satisfy the 30% rule.

    There are ways to complete the test without a load bank, says Dave Stymiest, PE, CHFM, FASHE, a senior consultant for Smith Seckman Reid, Inc., based in Nashville.

    For example, hospitals that want to skip the load bank may try to switch the facility over to generator power at night or on weekends.

    That can be problematic because there’s not enough bustle during the off hours to draw that 30% and give generators the realistic workout that they’d get when, for example, the ORs are in use, Stymiest says.

    Hospitals may be able get around that issue by turning on lights and similar equipment in empty ORs to simulate the power load of actual procedures.

    Some facilities will try generator testing during peak hours just to see if the emergency power supply system can sustain the rigors of a blackout.

    But Stymiest is “not a big fan” of this approach, he says. Imagine the risk of switching a surgical suite to faulty generator power in the midst of a patient procedure.

    “I think a better idea is to minimize the chance of something going wrong and adversely affecting patient safety,” Stymiest says. Hospitals can test generators during periods of low clinical activity as long as they realize that the stress on the generators will be less than during peak periods, he says.

  • Copyright © 2021. Hospital Safety Center.