Preventing generator failures with remote monitoring
Remote monitoring provides critical information that can help owners and operators prevent generator failures and ensure power system reliability.
Harold M. Jarrett, OmniMetrix LLC, Buford, Ga.
Most of us think of generator failure in terms of catastrophic events. We tend to picture pistons flying through the engine block, causing oil leaks and otherwise wreaking havoc. However, this is image is not at all consistent with what constitutes a failure to a genset owner. It is extremely unlikely that a genset owner will ever experience a major engine failure, but to that owner, a dead battery may be just as catastrophic when a utility outage occurs.
Typical causes of failure-to-start situations include the battery, fuel, block heaters, and ignored alarms. It is important to note that none of these failure modes are related to the genset brand; they are caused by lack of owner attention. In round numbers, 90% of failure-to-start events can be attributed to these four situations. This is where remote monitoring can make a significant difference. Even the most basic remote monitoring can easily detect these four problems. Then, prompt attention can prevent 90% of would-be emergency power system failures.
Batteries, fuel, and block heaters are consumables—they have limited lifetimes. Batteries have an installed life of about five years. Untreated fuel remains usable for one year. Expected lifetime of both can be shortened by customer site environmental factors such as temperature and humidity. The common weekly exercise is an important part of managing the battery and fuel. The genset's charging alternator delivers an important equalizing voltage to the battery when the engine is running. However, basic battery chargers provide only a constant float voltage. The fuel should be consumed, or else treated, within a year. Fuel consumption during the exercises helps achieve this. The genset’s operation, its incremental and cumulative run times, its fuel levels, and its battery charge voltages are fundamental to basic real-time remote monitoring, and remote diagnostics.
Installed generators occasionally have unexpected alarm conditions. About half of those alarms cannot be reproduced when the service technician arrives. These intermittent alarms can usually be attributed to transients from the utility, narrowly configured engine sensor settings, and sometimes just a glitch. Regardless of the cause, those alarms can make the genset nonfunctional to the owner.
A power outage during one of these transient alarms can result in the dreaded failure-to-start scenario. It is important to have the ability to diagnose alarms quickly and remotely to minimize the owner’s exposure time. In some generator brands, these alarms can be reset and the machine can be started and tested remotely. In the 50% of alarms that are simply transients, the owner’s standby power can be recovered quickly via this remote service attention. In the remaining 50%, the genset service provider can dispatch service with prior knowledge of the problem and appropriate replacement parts.
Transfer switches usually protect generators from utility voltage transients. However, their battery chargers and block heaters are not so lucky because they remain connected to the utility. These devices are common victims of voltage surges associated with power outages. In some cases, the generator control will produce an alarm indicating the charger failure. However, in most cases, the charger failure can be detected only through the diminishing battery voltage, which will commonly fall below engine cranking voltage within a day.
Figure 1 shows an actual example of a genset’s battery voltage drifting downward after a charger failure. Once the battery voltage drops below about 11.0 V, its capacity is severely diminished, and the constant load of the genset control drains it quickly. Detecting such an event is fundamental to remote monitoring.
Figure 1: This graph shows an actual example of a genset battery voltage drifting downward after the charger failed. Courtesy: OmniMetrix LLC
Fuel level monitoring becomes very important when the generator is remote and unattended. While virtually all generators produce a low fuel alarm from a float switch, this alarm is commonly set to actuate when a few hours of loaded run time remains before shutdown. While the absence of an alarm is a good thing, it often masks the reality of “almost low fuel,” where the available run time may be too short to replenish the fuel supply before shutdown. A better design point uses a variable level fuel gauge “sender.” This type of analog fuel level sender is able to indicate fuel level with measurements such as “Percent Full,” “Remaining Running Hours,” and “Remaining Gallons” or “Needed Gallons.” This detailed information—along with consumption trending—allows the owner to judge fuel supply urgency correctly.
Figure 2: This graph indicates poor power quality from the utility over a two-day period. Voltage fluctuations caused the generator to run excessively, consuming fuel unnecessarily. Courtesy: OmniMetrix LLC
Sometimes the expected fuel consumption is altered by unexpected running events caused by utility voltage. Obviously, power outages cause the generator to run. However, utility overvoltage is just as likely to occur and could cause a load transfer. Figure 2 shows an example of a location with particularly poor utility power quality, over a two day period, which caused the generator to run for more than 200 hours in a single month, completely consuming a tank of fuel. The benefits of remote monitoring in this case are twofold: it alerts the owner about the unplanned fuel consumption, and it clearly shows the cause of the problem—not the genset, but the utility.
Erratic utility voltage
Figure 3 shows the running events for this same site. None of these events were caused by utility outage; they were all caused by the erratic voltage tripping the generator into action. This is a very remote cell site. The traditional carrier operation center supervisory control and data acquisition system, which reports only power loss, would have ignored all of this activity. The site could have been a big box store, where the generator is hidden away behind the store and no one pays attention to the occasional operation of the engine.
Figure 3: This graph shows the generator running events for the same site that experienced the voltage fluctuations. During this two-week period, none of the events were caused by utility outages. Erratic voltage caused the generator to start because the system interpreted these events as power outages. Courtesy: OmniMetrix LLC
These examples show that the system reliability for the installed generator is very different from the engine-oriented measures of reliability. Since most standby generators run less than 100 hours per year, the expected tens of thousands of hours lifetime of bearings and mechanical parts translates into many years of expected run time before likely mechanical failure, vastly exceeding the expected lifetime of the consumables. So, regardless of the brand of generator, the emergency power system reliability is driven by factors outside of the engine. Monitoring can deliver critical information that can help prevent failures caused by these factors.
The reliability of an installed emergency power system is limited by the weakest link, and the failure of any of the links can result in a customer power catastrophe. These weak links are primary targets for remote monitoring. Prognostic detection and diagnostic analysis of their symptoms can be effective in dramatically reducing failure rates of emergency power systems, and simultaneously increase owner confidence.
Jarrett is the founder and vice president of OmniMetrix LLC, Buford, Ga., a company that provides remote monitoring solutions for generator-based emergency power systems.