Emergency power is one of the most critical lifelines in any facility, but it also offers a strong potential for surprise failures. Proper design can prevent or minimize problems, but maintenance and engineering managers also can help avoid these surprises through a focused preventive maintenance (PM) program.
Of course, some failures happen for reasons that front-line technicians can detect with routine testing. Many of these reasons relate to software that controls multiple-engine generator sets. For the most part, however, a strong testing and PM program will reduce the level of risk and raise the uptime percentage of a facility’s critical power system.
Generator and emergency-power failures can result from several common maintenance issues. These include the following:
• starting-system issues, batteries and cables
• bad fuel, fuel piping failures, and clogged fuel filters
• mechanical equipment issues
• light loading of generators
• control and software issues.
Managing these common failure areas allows managers to harden potential weak links in crafting the PM program.
Starting systems consist of batteries, starter motors and cables. Generator applications commonly use lead-acid batteries.
Negligent battery maintenance is the most common reason for failure of critical power systems. Visual inspections typically can pinpoint leaks, but managers also need to schedule in-depth testing of the cells and their charging ability to fully gauge the health of the system. By taking open-voltage and specific-gravity readings, front-line technicians can provide managers with a clearer picture of the state of a battery’s charge and locate a bad cell.
If the result of a battery open-voltage test or specific-gravity test turns out to be below manufacturer recommendations, it might indicate that a cell is going bad or that a battery was left in a state of discharge for too long. It is not sufficient to simply test the charging current to the group of batteries, however. The situation requires more reliable testing to evaluate each cell and replace those showing signs of degradation.
To help avoid surprise failures, technicians also should perform a visual inspection of all cables and connections, from the batteries and the starter motor to the battery charger. Technicians also should clean terminals and tighten connections to minimize electrical resistance in the circuit.
Fuel-related issues are the second most common failure point in emergency-generator systems. While this article references diesel fuel systems, some areas of the country also use natural-gas-driven generators.
Day tanks provide an immediate source of fuel for a generator in the event of a fuel pump failure or a line blockage from the main fuel storage. If not maintained, these tanks can be a weak link in the system. They need to be topped off and have the fuel moved and agitated, usually during generator testing. If technicians do not perform this task regularly, mold can grow in the fuel and clog fuel lines, starving the engine of fuel.
Mechanical issues are related to engine parts, cooling systems and other auxiliary equipment in the engine-generator set.
One of the most common failures relates to the jacket water heater or block heater that keeps the engine block warm to maintain oil temperature. Failure of this component can prevent the generator from starting. A good testing and inspection program can identify jacket water heater failure early so technicians can replace the heater to lower the risk of system failure.
Another mechanical failure relates to pumps in the system, including fuel pumps, oil pumps and those in installations where generator radiators are mounted remotely. These pumps can be weak links in a system and should receive monthly maintenance.
Many managers do not realize that generators, much like automobile engines, will last longer if they run at higher temperatures, which means loading the generator to at least 60 percent of rated load. Running diesel generators at loads lower than 60 percent can cause carbon build-up on internal engine components and can cause wet stacking.
This situation occurs when an engine-generator set operates below its optimum temperature. This condition allows unburned fuel to accumulate in exhaust system components and leads to degradation of fuel injectors, engine valves, turbochargers and other exhaust components, reducing efficiency of the engine.
National Fire Protection Association (NFPA) 110 stipulates exhaust-stack temperatures to prevent wet stacking based on generator size. For health care facilities, the Joint Commission of Accreditation for Hospital Organizations (JCAHO) stipulates the use of NFPA 110 for testing purposes.
For instance, the recommended stack temperature for a 1,250-kilowatt engine-generator is 700 degrees. The manufacturer of the engine generator set can give technicians the recommended stack temperature required to meet NFPA 110 requirements.
If it is not feasible to find enough building load to achieve a load of 60 percent or greater for testing, managers can buy or rent a portable resistive load bank for testing purposes to meet NFPA and JCAHO requirements.
Control systems for engine-generators, especially those that control multiple paralleled engine sets with automatic load shedding, can be very complex software programs. Inherent with any software program are system bugs that can be ferreted out only through successive testing against a stated and designed sequence of operations.
Typically, the commissioning process of the installation can address 95 percent of these issues, but latent programming issues can prevent generators from starting by sending false trouble signals.
In an emergency, plant operators can put the system in manual operation and chase down issues after handling the emergency. A solid monthly testing program will consciously seek sequence-of-operation errors and resolve them before an emergency arises.
Transfer switches often fail, so technicians should observe them during normal testing. NFPA 110 recommends testing procedures that include measuring millivolt drop levels across each pole. NFPA 110 also recommends that if any reading is greater than 25 percent of the average, technicians should inspect all poles.
Technicians also should take thermograph images of lug connections at the transfer switch. If the transfer switch has a bypass feature, the bypass should be operated, and millivolt drop levels should be measured across poles. Millivolt tests should not be performed on an energized transfer switch.
Circuit protective devices are critical pieces of the emergency power system to protect against overloads and faults. Electronic circuit breakers are more reliable than the thermal-magnetic type, but technicians must exercise breakers to prevent them from freezing up.
NFPA 110 recommends exercising circuit breakers under simulated overload every two years. Of course, this task can be difficult, given the need for the system to remain operational, and it requires diligent coordination.
These issues and recommendations are just a small part of a comprehensive PM program. Managers must recognize that emergency power systems are made up of multiple components, each of which requires some degree of maintenance.
To help fine-tune the program, managers need to keep clear and documented reports of failures, and review them regularly to find any historical trends. Managers also should keep an updated set of one-line power diagrams in main power distribution rooms. These diagrams can help with testing and ensure technicians have information at their fingertips to help them understand the entire distribution system.
By establishing a reliability-centered testing and PM program, managers can have greater confidence that when an emergency system needs to serve during a power outage, it will perform reliably.
Tom Leonidas Jr., P.E. — tleonidas@sparling.com — is vice president and a managing partner with Sparling, an electrical and technology firm with offices in Seattle and Portland.
