Fueling standby power options
Diesel benefits and drawbacks
Historically, diesel-fueled generators have dominated the market for larger industrial and commercial standby generators—those with outputs above 125 kW—by offering a significant first-cost advantage over similarly sized natural-gas units. However, this savings comes with its own premium: diesel designs require a steady fuel supply, which can become a problem in an emergency situation if roads are closed or a failed power grid shuts off fuel-pumping systems. As a result, the National Electrical Code’s Article 708, Critical Operations Power Systems, now requires emergency power-system operators to maintain a 72-hour fuel supply on-site, and the Joint Commission requires healthcare facilities to have 96-hour supply of fuel on hand.
But maintaining such an onsite diesel-fuel supply poses its own challenges. For example, the NFPA’s Standard 110, Standard for Emergency and Standby Power Systems, lists a number of requirements related to generator fuel systems, including:
NFPA’s concerns are focused on the understanding that the typical standby generator doesn’t run much in non-emergency situations. If the tank is sized for 72 hours of full load operation, it could take 22 years to get one fuel turn on the tank, assuming a 60% typical load level, weekly no-load exercising, and an average of 4 hours of outage per year. With such slow fuel turnover, contamination can become an issue.
The two most common fuel contaminants are water and biomass. Water enters the tank as humidity through the tank’s normal vent and condenses during the daily thermal cycle. Initially, moisture binders in the fuel capture and contain the moisture. However, as these binders become overloaded, the water drops to the bottom of the tank and begins accumulating. At some point, the moisture may be sucked into the diesel engine, potentially resulting in loss of power and lubrication, along with corrosion. Water also creates an environment that can support biomass growth at the water-fuel interface. When such microbes are pulled into the engine, they can clog the fuel filter, resulting in the engine losing power and shutting down. To minimize these effects, fuel tanks should have a well defined low point and the water should be removed monthly.
In addition to fuel contamination, fuel breakdown seems to be more prevalent with today’s low-sulfur fuels. The additional refining processes needed to remove the sulfur may be removing some of the fuel’s stability elements as well. The end result is increased fuel varnishing and gum formation. As diesel gets older, a fine sediment and gum forms, brought about by the reaction of diesel components with oxygen. Additives can help treat common fuel breakdown issues when integrated into a fuel-filtering preventive maintenance program, but, at some point, the fuel simply may need to be replaced. For example, BP recommends that fuel tanks should be emptied and cleaned at least once every 10 years and more frequently if there is a major fuel contamination.
Storing large amounts of diesel on-site also requires comprehensive handling and maintenance plans to ensure the system can deliver fuel reliably, while preventing any possibility for hazardous spills. Meeting both of these demands can boost system complexity, and result in the need for added monitoring and control capabilities. In an effort to minimize the negative aspects of on-site diesel fuel, systems that use simple sub-base (generator mounted) fuel tanks and limit on-site fuel to a more easily maintained quantity are generally preferred.
Natural gas alternatives
Spark ignited generators offer a number of advantages, including an extended run-time when they are paired with a reliable supply of natural gas. Through four Florida hurricanes in 2004 and the Northeast grid failure of 2003, the natural gas infrastructure was unaffected. Other benefits can include reduced permitting requirements, lower preventive maintenance costs, less risk of environmental contamination, and cleaner engine emissions.
Spark ignited (natural gas and liquid propane fueled) generators can compete with diesel models on cost in the automotive engine classification. While, historically, this has meant a maximum power output of 100 kW, various manufacturers have incorporated turbo-charging and optimized revolutions-per-minute technologies to boost this power range up to 125 kW, and even 150 kW, within the last few years. Above this output level, though, spark-ignited generators still pose capital cost premiums of 1.75 to 2.5 times the cost of comparable diesel generators. But manufacturers are developing new technologies to improve power outputs and extend the use of automotive-based engines to create new opportunities for natural gas installations.
Optimizing engine revolutions-per-minute ratings is one of the key technology shifts now broadening opportunities for spark-ignited engines. Fifty years ago, most generators operated at speeds below 900 rpm, but within the past 30 years, the diesel standby generator market has moved from 1,200 to 1,800 rpm as engine outputs have increased. The prime power natural gas driven generator market has also migrated from 900 to 1,200 rpm, with some recent offerings at 1,800 rpm.
These trends also have extended into the automotive-style, spark-ignited engines serving applications up to 150 kW. Historically operating at 1,800 rpm, current technology is optimizing these engines for operation at 2,300, 3,000, and 3,600 rpm. These greater operating speeds provide multiple advantages, including improved transient performance, less stress on engine bearings, increased power densities, and reduced capital cost. As manufacturers optimize operating speeds on automotive- and truck-derivative engines, the engines become more powerful and very cost-effective. The automotive engines (&150 kW) provide the greatest value at a significant discount to diesel engines, but optimized truck-derivative engines (&300 kW) also become cost feasible.
Packaging smaller generators into paralleled systems is another development that is helping spark-ignited engines become more cost-effective. To achieve this goal, manufacturers are using an integrated approach to generator paralleling, which connects the generators together and combines their output without using external equipment, such as switchgear. Parallel power solutions have always offered the standby generation marketplace significant advantages; however, implementing these solutions has been limited to mission critical applications and large kilowatt projects because of their cost and complexity. Eliminating switchgear removes much of the cost and complexity of earlier designs, so now three 500 kW gensets operating in parallel can offer the same output at similar cost as a single large 1,500 kW unit—but with the advantage of built-in redundancy. If for some reason the 1,500 kW unit does not operate, the facility is without backup power, but if one of the 500 kW units doesn’t run, the other gensets will supply two-thirds of the system’s normal output.
A third advance, bi-fuel generator sets, combines the power density and capital cost benefits of diesel engines with the extended runtime of natural gas. Using mass-produced diesel engines as prime movers, bi-fuel generators start up on diesel fuel in a normal manner, but as the generator picks up load, bi-fuel delivery systems introduce natural gas to the combustion air while reducing the amount of diesel fuel. Under full load conditions, bi-fuel generators operate on a ratio of 25% diesel and 75% natural gas, with no reduction in power.
At just a slight cost premium to diesel-only designs, bi-fuel generator sets offer several powerful advantages, including:
Because natural gas is the predominant fuel in these units, smaller diesel tanks are a viable option. With smaller fuel tanks, the risk of fuel going bad and the cost of fuel maintenance is significantly reduced. If the natural gas supply is interrupted for any reason, or if there is a fault in the bi-fuel system, the controls will automatically direct the unit back to 100% diesel, without interruption of operation.
Designers today have more options than ever when specifying standby power sources. Diesel is a highly reliable solution when integrated with a preventive maintenance program and refueling contingency plan, and natural gas has shown itself to be an extremely reliable fuel source with no interdependency on the electrical infrastructure. And, as the standby generator industry continues to adopt new technologies, choices likely will expand even further.
|Kirchner is industrial training manager with Generac Power Systems. Kirchner is an electrical engineer and beginning his career with Woodward Governor Co. designing hydro-turbine and plant control systems for the electric power industry. After leaving Woodward, he finalized his masters in business administration degree from the University of Wisconsin before joining Marathon Electric. There he preformed marketing and application engineering duties. In 1999, Kirchner joined Generac Power Systems.|
Sizing standby power for non-linear load applications
Matching a generator to an application’s load is, generally, a very straightforward process—the total load is measured or calculated, in kW, and the generator is selected based on the result. However, when the sizing involves loads that create significant amounts of harmonic currents (non-linear loads), the process can become very imprecise, often relying on rules of thumb and simplistic multipliers. Rules of thumb are valuable references, but non-linear generator sizing should incorporate some of the fundamentals of harmonic analysis.
Non-linear loads consume input power through their internal-switching power electronics. These switching devices consume 60/50 Hz current and produce other current frequencies—harmonics—as a byproduct. The harmonic currents vary, based on the type of switching devices, the number of switching devices (referenced as pulses) and the amount of harmonic filtration. The harmonic current’s amplitude usually is expressed as a percentage of the input 60/50 Hz current, while its frequency usually is expressed based on its multiple of the rated input frequency. Since non-linear loads produce many odd numbered multiples of the input frequency, harmonics are often referred to as a vectorial summation of the various frequencies. The common summations are THID (total harmonic current distortion) and THVD (total harmonic voltage distortion).
Not all non-linear loads are created equal. Typical three-phase equipment varies dramatically from 35% to less than 10% THID. So the first step in sizing a generator to a non-linear load is to recognize the load’s harmonic characteristics. A device with 35% distortion may need an alternator rated three times the size of the device, but a 10% distortion may require no alternator upsizing. As a general guide, loads with THID less than 12%, don’t require alternator upsizing.
So on what criteria should harmonic sizing be based? Traditionally, the market attempted to address this issue from a harmonic heating perspective, requiring alternators to be rated at class B and/or class F temperature rises. This approach may have worked with moderate applications, but when the load is predominately 6-pulse, unfiltered (35% THID) devices, the resulting voltage distortion will be extreme. IEEE 519, Recommended Practices and Requirements for Harmonic Control in Electric Power Systems, recommends that the maximum THVD level should be limited to 10% for a dedicated system. Keeping voltage distortion to a range of 10% to 12% is a good target—application experience tells us that voltage distortion in the 15% to 18% range generally results in application problems, while limiting distortion levels to less than 10%, through alternator sizing, may be cost prohibitive and best done through adding harmonic filters to the loads.
Moving from the harmonic current of the device to a resulting system voltage distortion is as simple as applying Ohm’s law. Ohm’s law predicts that when current flows through an impedance, a corresponding voltage will result. Though the math of harmonics doesn’t allow the equation to be as simple, the same effect occurs with harmonics. When load-produced harmonic current flows through the source impedance of the alternator, a corresponding voltage distortion is created. To control the distortion to 10%, either the harmonic currents need to be reduced or the alternator’s source impedance needs to be reduced.
The source impedance of interest is the alternator’s sub-transient reactance (x”d). The market occasionally requests alternators with subtransient reactance less than 12%, as a method of securing harmonic performance. Though this is a good starting point, if the non-linear load is a typical 6-pulse, unfiltered device, the equivalent reactance needs to be closer to 6%. This source impedance is usually given in percentage values, typically ranging from 10% to 18% for three-phase generators, and 14% to 24% for single-phase equipment. Alternator impedances are much higher on small machines and single phase equipment, so there is a greater need for proper harmonic analysis with these devices. It is common for small single phase alternators to be sized five times the non-linear load level.
One common tactic for reducing generator-set impedance is to upsize the alternator, because larger alternators have lower source impedance. However, because alternator reactances are in a percentage system, it is important to recognize that the reactance number is relative to the base kVA of the machine. This means that specifying an alternator with a larger base kVA effectively lowers the absolute value of the machine’s impedance.
As this discussion illustrates, sizing generator applications based on harmonics can be challenging. Various tools exist to aid in this analysis. When using generator sizing tools, scrutinize the effectiveness of the non-linear load sizing based on the tool’s inputs and outputs. If the sizing tool uses inputs of harmonic currents and alternator reactance and the output includes a calculated voltage distortion, the tool is performing a harmonic analysis.