Engineering of Water Systems

Standby and Backup Systems: Sizing Generators

By Ed Butts, PE, CPI

The Water Works column in the January 2025 issue of Water Well Journal began a four-part series on generators used for standby and backup water and wastewater service.

Part 1 included an overview of the theory and function of AC and DC generators. This month, as Part 2, we will review the general considerations for generator sizing and selection. The series will wrap up with Part 3 in July dedicated to generator sizing and selection for electric motor service and Part 4 in October providing real-world examples of generator sizing.

Generator Sizing and Selection

There is no single correct sizing procedure. Several methods are used, and when combined with proper judgement and design practices, should result in an appropriately and economically sized generator.

Figure 1a. Simplified three-phase AC generator.

This includes remembering to consider all current loads, load growth potential, load factor, seasonality impacts, and effects of starting motors, including cycling (intermittent) operation common to water system applications.

Always check with the local electrical or building inspection department (the Authority Having Jurisdiction [AHJ]) to confirm which NEC code cycle and local codes will affect the design and installation.

Figure 1b. Voltage output of a single- and three-phase generator.

Equipment, such as most domestic water system pumps, use power on a cycling basis and are generally considered as intermittent loads. All loads that follow an intermittent load must consider the intermittent load as part of its total. These increase starting (skVA) requirements may mean a larger generator is needed to account for the intermittent loads.

Never simply add the full load amps from each motor when sizing a generator. Always convert the motor’s nameplate amps to kW first and then factor the starting requirements in sKVA method of motor starting and load factor (number of motors running at any one time). Finally, add all resistive loads to determine the required generator size.

Figure 1c. Fundamental single-phase AC generator.

Adding amps without properly accounting for the starting and running power factor or mixing voltages on an AC system may result in improper sizing of the generator since power factors for motor loads can vary widely. When motors start, they create a momentarily higher current surge that increases the load on the generator, which creates a voltage dip.

A properly sized generator set will support the high starting kVA requirements of a motor while maintaining adequate output voltage for the motor. Thus, it can develop adequate torque to accelerate the load to rated speed.

Most standby gensets use synchronous generators with exciters; many are available with permanent magnet generator (PMG) excitation systems. The PMG provides excitation power independent of them generator terminal voltage. As such, it can maintain full excitation even during transient loading, such as motor starting.

Full excitation power results in a less extensive voltage dip and improved recovery times. After preliminarily selecting am generator, reference the generator’s surge capability and verify that the voltage dip (ΔV) is not excessive for the application. The larger of voltage dip a generator set can tolerate, the smaller and more economical the generator set can be.

Most water system applications with full voltage start motors should be limited to around a 15% ΔV maximum. However, voltage dip with residential and non-critical applications are often limited to a 30% maximum voltage dip.

Frequency dips are related to the size of load being connected. A 10% to 15% maximum frequency dip is the standard limit and a frequency dip above 20% may cause the engine difficulty in recovering. Frequency dips are normally more tightly controlled than voltage dips because they are typically connected to more sensitive electrical equipment.

Always check with the local electrical or building inspection department to confirm which NEC code cycle and local codes will affect the design and installation.

Nonlinear loads, such as variable frequency drives and switching power supplies, also require specific attention because the SCR switching causes voltage and current waveform distortion and harmonics. The harmonics generate additional heat in the generator windings. Therefore, the generator may need to be upsized to accommodate this potential.

In addition, some applications utilize an uninterruptible power supply (UPS) to back up critical loads or welders that each draw intermittent and high generator loads. Therefore, depending on the application, UPSs, variable frequency drives, welders, and all other non-linear loads require guidance and sizing by the genset manufacturer.

Motors require special consideration with generators. To ensure adequate torque is available to overcome the inertia needed to start a motor, particularly submersible and vertical turbine pump motors, a 15% to 25% voltage dip is a typical recommended limit for most motors and industrial motor controls. However, sensitive electronic loads, such as variable frequency drives, may be specified for much lower voltage dips.

Typically, the maximum frequency dip for motor starting is limited to about 25%, but modern monitoring equipment is often restricting this to tighter margins—down to 5% or less. Load “steps” are used to determine the progressive demand on the genset during each successive event of adding load to the unit, including motor starting, adding resistive load, or a combination of both.

Most manufacturer generator selection programs are designed to consider each of these variables during the selection of a generator. The key to choosing the right rating and generator set is to understand the application in detail. That means not only knowing the duty but answering the following:

• If AC power is planned, define if the generator is to be single- or three-phase and the output voltages.
• What is the maximum and minimum required running load?
• Are operating motors involved? If so, how many are there, what horsepower, starting and running power factor and efficiency, and how many must be started simultaneously or in steps? What is the allowable motor starting voltage and frequency drop and the required or desired motor starting method?
• How many hours per day or year will the generator set likely run?
• Does the generator provide emergency power, legally required standby, or optional standby power?
• What power rating is required: prime, standby, or continuous?
• Are there any special or unique voltage or frequency requirements?
• What is the average load grouping, demand factor, and diversity?
• Is the proposed genset intended to operate in a stationary position or be portable?
• What type of engine and fuel type is required?
• What is the acceptable or allowable percentage of operating voltage and frequency dip?
• What is the acceptable duration of the voltage and frequency dip recovery time between steps?
• How many steps? What percent of a load step and type of load is to be connected or added during each step?

Many people believe smaller generators can be used for standby electric power because they are not running all the time. This is not only a myth but can be dangerous to systems where life safety operation is vital.

Unfortunately, generator undersizing is one of the most common mistakes committed by designers or buyers. Not only does it risk damaging the generator itself but can also damage other assets connected to it as well as create hazardous situations and limit overall productivity of the unit and the entity or water system relying on it. If nothing else, the key to remember with generator sizing is that a little more is always better than a little less.

Figure 2. Three-phase generator principle.

Single- and Three-Phase Power Generation

Single-phase and three-phase induction generators fundamentally operate in the same manner, with the difference being the use of a single winding in a single-phase unit and three windings in a three-phase generator (Figure 1a).

Both units produce voltage that corresponds to the rotor’s north-south position relative to the windings (Figures 1b and 1c). When three coils are connected rigidly together and set 120° electrically apart but rotate about the same axis in a uniform magnetic field, the induced EMF (indicated by an “E” in Figure 2b and Figure 2c) in each of them will display a resultant phase difference of 120° or 2.09 radians.

Figure 3. Power configurations for wye (Y) and delta (Δ).

Consider the three identical coils: a1a2, b1b2, and c1c2, as shown in Figure 2a, each mounted on the same axis but displaced from each other by 120° and rotating in a counterclockwise direction within a bipolar magnetic field.

Here, a1, b1, and c1 are regarded as the start or beginning terminals and a2, b2, and c2 are the finish or ending terminals of the three coils. When the coil a1a2 is in position AB, the induced EMF in this coil is at zero but is increasing in a positive direction; the coil b1b2 is lagging 120° behind coil a1a2 so that the EMF induced in this coil is approaching its maximum negative value; and coil c1c2 is 240° lagging behind coil a1a2 so that the EMF induced in the coil has passed beyond its positive maximum value and is now decreasing.

Since each coil, being identical, has an equal number of turns and is wound with wire of the same type and size, the induced EMFs in each of the coils are also the same magnitude. The induced EMF in each coil is also the same frequency and waveform as a sinusoidal wave but displaced from each other by 2.09 radians or 120°, as illustrated in Figure 2b by the waveforms.

Phase sequence is the order or sequence in which the currents or voltages in different phases attain their maximum values, each one following the other. The three coils, a1a2, b1b2, and c1c2, are rotating in a counterclockwise direction. Since the coil b1b2 lags 120° behind coil a1a2 and coil c1c2 lags 240° behind coil a1a2, the first coil, a1a2, attains a maximum or peak value of induced EMF, then the next coil, b1b2, attains a maximum or peak value of induced EMF, and when the coils rotate further another 120° (electrical), coil c1c2 attains its peak value of induced EMF when the coils rotate through 240° of total electrical shift as demonstrated in Figure 2c.

Since the induced EMFs in the three coils, labeled as a1a2, b1b2, and c1c2, attain maximum values in order of: a, b, c, the phase sequence is also: a, b, c. If the direction of the rotation of the coils is reversed to be clockwise, the three-phase EMFs would now attain their maximum values in the different order of: a, c, b. Thus, the phase sequence would now also be a, c, b.

Since the coils can be rotated in either a clockwise or counterclockwise direction, there can be only two possible phase sequences. The knowledge of phase sequence of a three-phase power supply is essential for making the proper connections for alternators and transformers to operate them in parallel.

The direction of rotation of an induction motor also depends on the phase sequence. This is because if the phase sequence is reversed by interchanging any two terminals of the three-phase power supply, the motor will also rotate in the opposite direction. Depending on the application and type of driven machine, this can result in no damage whatsoever up to a potentially disastrous outcome—up to unscrewing of lineshaft
on a vertical turbine pump or unscrewing of pipe joints and separation of drop cable on a submersible pump.

In general, the phase sequence of the voltage applied to the load is defined by the order in which the three phase lines are connected. In the case of three-phase unbalanced loads, the effect of reversing the phase sequence will cause a completely different set of values of the line currents. Therefore, while working on all such systems, it is essential to specify the phase sequence clearly and to connect each phase in the proper order to avoid unnecessary confusion and potential equipment damage.

There are two types of connections used for three-phase power supplies, delta (Δ) and wye (Y) (Figure 3). The connection shown in Figure 2c is an example of a wye connection. The wye phase connection has a neutral point, often connected to ground, but the delta configuration does not have a neutral point.

The wye configuration is used with three-phase loads or lineto- line loads while delta connections are used with single-phase and three-phase loads. Common wye connection voltages include 120/208 VAC and 277/480 VAC. Delta connections are found in rural communities where three-phase power is not available. Common delta voltages include 120/240 VAC, single- and three-phase, and 480 VAC, three-phase.

A generator can lose up to half of its rating if a delta configuration is hooked to a three-phase system. It is most common for generators to have 12 or six leads, with the 12-lead generator the most popular type in the United States. A 12-lead, wye connected, 480 VAC generator set is usually required for larger pump motors and other large power demands.

The use of 240 VAC power is typical for smaller single- or three-phase motors, up to 15 HP in single-phase, and 30 HP in three-phase. 120/240 VAC is commonly used for single-phase motors and other single-phase loads.

Understanding Generator Ratings and Terms

AC generators are typically sized using power output as the determining factor, using units of kilowatts (kW) or kilovolt- amperes (KVA). They are available in single- or three-phase windings with wye (Y) and/or delta (Δ) configurations and multiple voltages, from 120 VAC to over 16,000 VAC.

The generator is sized to supply the kVA needed at startup and the kW during normal running operation. It also provides voltage control using a brushless exciter and voltage regulator.

Together, the engine and generator provide the energy necessary to supply electrical loads in many different applications encountered in modern society. The generator set must be able to supply the starting and running electrical load and must be able to pick up and start all motor loads and low power factor loads and recover without excessive voltage dip or extended recovery time.

It is also necessary to determine the load factor or average power consumption of the generator set. This is typically defined as the load (kW) times (hours while under that particular load) divided by the total running time. When this load factor or average power is taken into consideration with peak demand requirements and the other operating parameters mentioned above, the overall electrical rating of the genset can be determined.

Other items to consider include the unique installation, ambient, and site requirements of the project. These will help to determine the physical configuration of the overall system.

Current output is a function of the delivered power in kW and directly related to the connected voltage. A complete description of the relationship between watts, voltage, and current will be outlined in Part 3 in the July 2025 issue of WWJ.

Standby power-rated generators are the most commonly used generator sets. Their primary application is to supply emergency power for a limited duration during a temporary or short-term power outage. With standby-rated generators, there is generally no overload or reserve capacity built into the units.

It is important to note that under no circumstances should standby-rated generators be allowed to run in conjunction with a public utility power source.

ISO-8528-1:2005 is an industry standard for performance parameters in on-site power applications. The ratings outlined in ISO-8528-1:2005 define basic generator set rating categories based on three segments: continuous power, prime power, and limited-time running time.

In each category, a generator set’s rating is determined by its maximum allowable power output in relation to running time and load profile. Misapplication of these ratings can jeopardize the longevity of the generator set, void manufacturer’s warranties, and put the set at risk for failure.

______________________________________________

This concludes Part 2 of a four-part series on generator sizing and selection for standby and backup service. In July, we will continue with a review of the various starting methods for electric motor applications.

Until then, keep them pumping!

Ed Butts, PE, CPI, is the chief engineer at 4B Engineering & Consulting, Salem, Oregon. He has more than 40 years of experience in the water well business, specializing in engineering and business management. He can be reached at epbpe@juno.com.