Power When It Counts
Strategic testing of emergency-power systems protects hospitals and patients from dangers created by disruptions
Hospital maintenance and engineering departments must have programs in place to monitor and test their facilities’ emergency power systems. With the recent release of the Sentinel Event Alert 37, the bar is higher for all health care organizations accredited by the Joint Com-mission for the Accreditation of Healthcare Organizations (JCAHO).
Before the alert, the recordkeeping for many hospitals’ testing programs consisted mainly of generator loading and operating parameters, test times, and verification of transfer switch operation. Now managers must consider other issues, including test-related interactions with HVAC, vertical-transportation, and clinical systems and equipment.
Among the additional issues that managers must consider are these.
Design to Facilitate Testing
Facility managers should ensure that design engineers for new construction and renovation projects understand the periodic testing requirements and that they factor those activities into new-project designs.
One example includes using uninterruptible power supplies (UPS) for building automation systems and other control systems so periodic load transfers do not cause a system to reset. Where a UPS is used, it should be included in the facility’s maintenance management system so that it continues to operate reliably.
Another example includes telling design engineers the way that technicians conduct periodic tests, as well as steps they can take to make testing less labor intensive.
Commissioning and Acceptance Testing
Managers should arrange to commission new emergency power systems and major system modifications. These important systems need this extra level of job site quality assurance. Assuring the systems are constructed in full accordance with the design intent also will reduce test-related problems.
One recommendation is the four-hour NFPA 110 Initial Acceptance Test, witnessed by the Authority Having Jurisdiction. This test is generally the only way technicians can test the entire emergency-power system to ensure all appropriate loads are properly connected to the system and are able to run from that system. This goal is accomplished by the test’s design.
The first two hours simulate a primary utility power failure, or full normal power outage, to verify all facility loads required to be on emergency power are, in fact, powered from and able to run on that system. Also, during this first two-hour load test, paralleled systems must use and test all paralleled generators that are needed to operate simultaneously.
The second two hours is a full-load test. This test must apply a 100 percent rated load in one load step using a load bank with or without building loads. This portion of the test is intended to prove that the newly installed system can power a rated load.
A proactive testing program for emergency power involves knowing the load profile of the emergency power system. Managers can determine this by annually monitoring the transfer switch loading — or the load side of the switching apparatus in systems that do not use transfer switches.
NFPA 70 220.87, Determining Existing Loads, requires one year of maximum demand data. If one year of demand data is not available, the NEC requires technicians to use a recording ammeter or power meter to document 15-minute demands for 30 days, with seasonal and periodic loads factored into the 30-day readings.
Loading for Emergencies
Such loading is likely to be greater than the loading during monthly tests because of the test time, which probably was chosen because it is a period of low clinical activity, when the operating rooms and radiology procedure rooms generally are not in use. As a result, the monthly test load will not reflect the avoided clinical activity.
During a normal power outage, the emergency power system loading will reflect the time the outage occurs and is likely to include more clinical activity. The loading during an emergency might include extra fire-alarm system loading, smoke-control systems, and the fire pump load, if the emergency is internal. It also might include an extra clinical load, due to an emergency-department surge that depends upon the nature of the exterior emergency.
Monthly Load Tests
Although monthly load tests are functional tests of emergency power generation and distribution equipment, they also represent the best time to monitor and consider more than just the engine parameters and the generator set loading.
These monthly load tests are an opportunity to train front-line maintenance technicians, as well as clinical personnel, in the nuances of the electrical environment of care. This area of training should cover mechanical-system responses, other building-system responses, and clinical equipment responses to power transfers.
Analyzing Results and Failures
The emergency power testing program also includes a monthly review and analysis of the test’s results and its failures. Failures will occur during monthly tests because some of the equipment that is normally off is operating.
These failures are not bad news. Rather, they are opportunities. Most test failures would have occurred anyway during the next unanticipated normal power outage.
A failure during a scheduled monthly test — under carefully controlled conditions with heightened awareness — is much more benign than a failure that occurs during an unanticipated outage. These failures actually allow the facility to avoid such future problems.
Trend Analysis of Results
Managers can use spreadsheet or database software as a management tool to assist with trend analysis of test results. Keywords are a useful tool for this purpose, as explained more fully with examples in the management monograph from the American Society for Healthcare Engineering (ASHE) discussed in the accompanying article.
Trend analysis facilitates investigating and resolving training and systemic issues revealed with the monthly testing.
Extended-run Load Tests
The new Element of Performance 5 for JCAHO EC.7.40 requires a four-hour load test every three years. JCAHO requires only a 30 percent generator loading for this test, and it allows either a load bank — static — or actual loads — dynamic.
But due diligence considerations for systems with actual loads greater than 30 percent suggest the loading should be at least to the system loading, rather than to the JCAHO minimum. Managers also might consider performing the NFPA 110 extended-run load test, which will meet the criteria for the JCAHO extended-run load test with dynamic loads, as long as system loading exceeds 30 percent.
Resistance to Power Transfers
Many health care facility managers still must deal with clinical resistance to power transfers, despite the clear requirements for monthly operation of all transfer devices. It is understandable that a manager would not want to place patients at risk, for example, by transferring an operating-room critical branch switch with surgery in process or by transferring the switch serving rooftop helipads with an incoming medical emergency flight.
But it is necessary to transfer all skipped transfer switches sometime later during that same 20-40 day period. Many managers schedule backup dates and times every month to test skipped transfer switches. These backup times might occur when managers anticipate little or no clinical activity.
Refusing to operate transfer switches at all due to clinical concerns is actually doing a disservice to patient care because that approach masks potential latent defects that are likely to reappear during real normal power outages. Facilities should get to the real technical reason behind the clinical resistance and deal with the problem, not just the symptom. Just as patients want clinicians to cure their illnesses, not just treat their symptoms, the facility managers should strive to find and fix the problems behind the clinical resistance. The ASHE monograph offers suggestions for dealing with these issues.
Written test procedures are helpful for many reasons. They can give managers control over the test process and allow them to take into account variations in equipment, they require technicians to take responsibility for performing all required tasks, and they reduce the chances of technician mistakes.
Since testing is intended to catch incipient failures before the next utility outage, written procedures with spaces for initials and time entries also provide written documentation of the actions taken during the test. This precaution can be helpful in researching failures.
Finally, written test procedures will better enable managers to explore potential trends, as well as source documentation for later analyses.
David Stymiest, P.E., CHFM, SASHE, is a member of the compliance and facility management team with Smith Seckman Reid, an engineering design and facility consulting firm with headquarters in Nashville, Tenn. He also is responsible for regulatory compliance, facilities engineering and management consulting for the firm’s health care clients.
Editor’s note: Although the author is chair of the NFPA Technical Committee on Emergency Power Supplies, which is responsible for NFPA 110 and 111, views and opinions expressed here are the author’s. They shall not be considered the official position of NFPA or any of its technical committees. They also shall not be considered to be, nor be relied upon as, a formal interpretation. Readers can refer to the entire text of all referenced documents. NFPA members can obtain NFPA staff interpretations at www.nfpa.org.
Guidance on Testing Tactics
One resource for engineering and maintenance managers trying to get a handle on the new approaches to electrical system testing are publications from the American Society for Healthcare Engineering (ASHE).
The recent Sentinel Event Alert 37 from the Joint Commission on Accreditation of Healthcare Organizations references a 2003 ASHE management monograph on testing hospital emergency-power systems, as the author of both documents. I recommend that managers obtain the new expanded version, a 2006 update, Managing Hospital Emergency Power Systems – Testing, Operation and Maintenance.
This version offers a more comprehensive treatment of the issues raised in Sentinel Event Alert 37. The updated document is available free from ASHE’s web site, www.ashe.org.
— David Stymiest