Ensuring emergency power performance
Several rivers in the Midwest swelled well beyond their banks in 2008, causing multiple levee failures and widespread flooding. Homes and businesses became islands surrounded by raging waters. When the flooding knocked out the municipal power supply, many relied on emergency generators to provide electricity.
Businesses and industries commonly rely on emergency generators to provide power in the event of utility outages. Emergency generators and emergency generator systems provide backup power for everything from hospitals to manufacturing plants. But emergency power generators sometimes provide a false sense of security. If proper precautions are not taken to ensure their reliability, emergency power might not be available.
The size and voltage of emergency generators varies greatly. The emergency power needs of a typical homeowner pale when compared to those of a business or large industry. For business and industry, a typical emergency power supply system (EPSS) has one or more engine-generator sets with battery or air starting equipment, a local fuel “day” tank and remote fuel bulk storage tank, and an automatic transfer switch (ATS) or generator switchgear. The output is distributed by a power distribution system consisting of panelboards and feeder or branch circuits, which connect to the various emergency loads.
Thorough consideration should be given to the design of an EPSS, especially when it serves facilities where the physical protection of the components is especially challenging. There are multiple issues to take into account, yet several require only a common-sense approach.
Codes and standards
The codes and standards for EPSSs are well researched and detailed due to the importance of emergency power in critical facilities. Examples include NFPA 70 National Electrical Code Articles 700, 701, and 702 on emergency and standby systems; NFPA 30, 99, 101, and 110; and relevant state and local codes. However, much of the content of these standards deals with EPSS internal operation and how the system is expected to respond to the loss of the normal power service, which is the primary function of an EPSS. The standards have few requirements for physically protecting such systems from external threats, such as fire, wind, or flooding.
It is left to the engineer to identify all threats (physical and electrical) to each project and address them with facilities or operating procedures. Some threats, such as hurricanes, are unique to certain geographical areas and addressed by special local code requirements. Where code requirements are not clear, and the physical threats to the EPSS are obvious, the engineer must make the case for a higher degree of physical protection.
Engineers deal with the requirements for EPSS on a facility-by-facility basis. High-profile facilities, such as hospitals, police stations, fire stations, and public shelters, are considered essential. However, there may be other facilities in a region that do not have a public profile but whose failure can lead to devastating consequences. One example is a drainage pump station. The operational failure of a drainage pump station can cause extensive flooding, making its proper operation even more critical than that of more obvious, high-profile facilities. In the end, these necessary facilities cannot be allowed to fail for any reason. Accordingly, an equal level of performance is expected from their EPSS.
An emergency power source cannot perform successfully unless each system component functions properly. While attention often is given to protecting the engine-generator set, components like the fuel system, ATS, and the emergency power panel are not given the same degree of care. As a result, the emergency power source may fail.
For example, if the engine-generator set for a hospital is installed on the roof to avoid being flooded, but the automatic transfer system and emergency power panels are placed in the basement, then the system is likely to fail when floodwater rises in the basement, regardless of the reliability of the engine-generator.
Many items, while seemingly unimportant, can jeopardize the operation of the EPSS. Consider convenience receptacles or light fixtures at a low elevation after a flood. They may be served by overcurrent devices with poor selective coordination that can trip circuit breakers farther up in the system. As a result, portions of the system far removed from the receptacles or lights can be disabled.
A good rule of thumb in EPSS design is that Murphy's Law always prevails: “Anything that can go wrong will go wrong.”
In the conventional testing of an EPSS system in accordance with NFPA 110, the entire system is tested to see what happens when the utility or normal power source fails. This is most easily done by tripping the main utility circuit breaker serving the facility or serving normal input to the ATS. The main issue with this “cold turkey” test often is convincing the emergency load users to allow such an analysis to take place. Assuming that this is accomplished, the test can be scheduled, and the successful completion should provide a high level of confidence in the reliability of the system.
Testing the EPSS in the presence of physical threats to the components is much more difficult. A Category 5 hurricane or a 200-year flood (or both at the same time) cannot be conjured to test how the system reacts to natural disasters. It is necessary to anticipate what the worst combination of threats does to the power supply. For example, an engine-generator with a weatherproof enclosure rated for 150 mph winds and suitable for a rainstorm may not be suitable for a 150 mph wind-driven rainstorm. It certainly will not be suitable for a 150 mph wind-driven 2x4 that can penetrate a brick wall. In any case, testing the worst-case scenario isn't practical.
Some EPSS tests can be done over a short period of time and provide confidence that the entire system will work for much longer. But occasionally, the EPSS should be tested for the full time period and electrical load that it will experience in a real natural disaster. It is possible that the system works well for four hr, however, when subjected to a 12-hr test, it suddenly fails after just six hr because the fuel day tank runs dry. Closer inspection shows that the fuel transfer pump that automatically fills the day tank is powered from the normal power system, and not from the EPSS. When tested only for four hr, the tank was filled by the normal power source when the test was finished, and this fatal flaw in the system was not discovered.
Most small systems are configured with an engine-generator, an ATS, and an emergency power panel. Two viable sources of power are required, and this seems to be a cost-effective emergency power solution because the utility is usually highly reliable. If the utility fails, the ATS starts the engine-generator, which appears to supply power until the end of the contingency.
If the facility has to perform without fail during a natural disaster, there may be only one viable power source. When a regional windstorm occurs, as happens during a tropical storm or hurricane, the utility can de-energize the service so the EPSS really has only one viable source of power. It is hoped that the one source will be reliable through the entire event.
Depending on the criticality of the facility, at least two—and maybe more—viable power sources are needed. For example, pump stations in Louisiana's Jefferson Parish are being upgraded to have two viable power sources, not counting the utility. An elevated safe room on each site, which has its own small engine-generator set and ATS, will be reconnected from a dedicated utility service to be served from the EPSS of the pump station. Then the safe room will have three viable power supplies.
If engineers choose to have two viable power supplies in addition to the utility, it's essential that they are independent of each other and not likely to be disabled for the same reason. Two engine-generators served from the same fuel day tank or located immediately adjacent to each other are not truly independent power sources. One failure scenario can disable both engine-generators at the same time.
Preparing for the worst
During Hurricane Katrina, pump stations in Jefferson Parish lost large roof sections and in one case an overhead door, resulting in a large volume of high-speed wind-blown rain entering the building. The interior EPSS electrical equipment had NEMA 1 indoor enclosures because the engineers had not expected the building enclosure to fail. A cost-effective solution would have been to provide NEMA 3R enclosures, which are generally available for most electrical equipment from large switchgear line-ups to small disconnect switches. (See sidebar, page 21.)
Electrical equipment often can be fire-proofed for very little cost by simply installing it at a higher elevation. Unfortunately, at whatever elevation the equipment is installed, there is always a higher elevation at which floodwater can disable the equipment. In the end, the best approach is to install the EPSS equipment at an elevation that is sufficiently higher than the critical loads it serves. If precautions are not taken to ensure its reliability, an EPSS might represent a false sense of security rather than provide resiliency against natural and manmade catastrophes.
Read " Cleaning up after Hurricane Ike " from the June issue of Pure Power to view more information about emergency power performance and the effects of natural disasters on major urban areas.
McAllister is a project manager and principal electrical engineer with Stanley Consultants. He has nearly 40 years of professional experience in the design of building facilities and infrastructure projects with emergency power supplies.
Case in point: Storm-proofing Jefferson Parish pump stations
On Aug. 29, 2005, Hurricane Katrina hit Louisiana with winds reaching 175 mph and massive storm surges, breaching many levees as it powered its way through New Orleans. Much of the city was eventually flooded, and floodwaters lingered for weeks before finally draining (or being pumped) back into the Gulf of Mexico.
Stanley Consultants has provided engineering and architectural services to the U.S. Army Corps of Engineers for the recovery of hurricane/flood protection facilities in Jefferson Parish. The facilities were greatly affected by Katrina but were not damaged to the extent of those in New Orleans or areas farther east. Objectives of the project were to storm-proof pump station buildings and equipment, as well as provide an elevated safe room the pump station can operated up to and immediately after a storm.
The work involves design of storm-proofing measures for large drainage pump stations. Following are eight ways to improve sustainability of a pump station operation.
Strengthen pump station building structures
Elevate pump drives, motors, switchgear, and other vulnerable equipment
Provide backup power for at least one pump, emergency lighting, miscellaneous electrical loads, and redundant power generation
Flood-proof cooling water wells and fuel tanks
Electrical: Flood-proof switchgear, transformers, starting reactors, rheostats, etc.
Mechanical: Flood-proof the pump motor pit drainage system; adapt gear drives for raised motors; use remote control for sluice gates operators; and raise fuel tanks and/or fuel tank vents
Provide additional protected fuel storage on-site for extended operation following a storm
Provide structural flood-proofing consisting of raised door sill heights, enhanced structural roof integrity, and roll-up steel shutters over windows and doors.
13 tips for emergency power performance
The following questions form a basic checklist for evaluating an emergency power supply system (EPSS) against physical threats. These should be considered during the design and commissioning of a new system or when evaluating the vulnerabilities of an existing emergency power system.
Does the EPSS have enough viable power supplies for a worst-case disaster? What disasters will the EPSS be protecting against?
Are the viable power supplies essentially independent or can they be disabled for the same reason?
Is there enough fuel stored or available from a supplier for a worst-case event? In case of a disaster, does the fuel supplier have the capability and commitment to provide fuel needs?
Has the emergency power system been tested for a time period that is as long as the worst-case disaster scenario?
Has the system been subjected to a “cold turkey” utility outage test?
Has it been thoroughly tested since the last time modifications were made?
Are the indoor components of the EPSS protected from elements if the building enclosure fails?
Are the outdoor components protected from wind-blown missiles? At what speed?
If engine-generators have compressed air starting, are redundant engine and motor-driven air compressors available?
Does the system meet the requirements of NFPA: 70 National Electrical Code, NFPA 110: Standard for Emergency and Standby Power Systems, and the local and state codes?
Is the automatic transfer switch in working order?
If the facility floods, is the EPSS likely to stay in service at least as long as the emergency loads?
Where does the facility fall in the priority scheme of the utility? (For example, a hospital is the first load to which a utility typically restores power.) How close to the power source is the facility?