Coal, O&M

Stand By for Parallel Power

Issue 4 and Volume 104.

Facilities with Critical Electric Power Needs Often Install Standby Generators to Supply Power when normal utility service is interrupted. Critical needs include air conditioning, lighting, computer networks, security systems and communications. Even facilities with battery-backed uninterruptible power supplies (UPS) usually need standby generators to serve loads for extended duration outages.

Click here to enlarge image

In the simplest of installations, a single standby generator set is started up and switched on to the load by a transfer switch when there is a utility outage, and it is shut down when utility power is restored. However, there are a number of circumstances in which it is beneficial to have multiple standby generator sets running in parallel to supply emergency electrical loads. Further, depending on local electrical rates and the utility’s distributed generation policies, there may be economic incentives for having on-site generators running in parallel with normal utility power for load sharing.

Parallel Power

Parallel power is the operation of two or more sources of AC electrical power whose output leads are connected to a common load. The power sources connect to a common busbar and function electrically as a single source of power. To be paralleled, two sources of power must have the same number of phases and matching voltage and frequency. The generator system may, in the case of interruptible applications, be paralleled with power from a utility source (see figure).

Two 1,250 kW gensets comprise the PowerCommand paralleling system used for standby and interruptible service at H.B. Fullers’ adhesives facility in Minnesota.
Click here to enlarge image

A facility might have any of several reasons for operating two or more generator sets in parallel on a common bus, rather than operating a single, larger generator set. Paralleling generator sets can be more economical than running a single larger generator set. For example, an existing distribution system may not lend itself to being split into separate sections handled by separate non-parallel units. It may be cheaper to parallel generator sets and run the entire distribution system from the paralleled sets than to split the distribution system.

The original equipment costs per kW are also lower on smaller, 1800 RPM generator sets than on larger, lower RPM sets. Costs are also lower for the breakers and other system components for smaller sets. These lower initial costs for generating equipment can be significant factors in selecting the appropriate generator system for a site.

PowerCommand’s digital master control panel.
Click here to enlarge image

Paralleling also permits closer matching of the power produced to the actual loads that are drawing power. For example, it may be possible to operate a single generator set when loads are light. When loads increase, other generator sets in the paralleled system can be added. This type of operation, known as load demand mode, is often used in prime power or during long power outages, and saves fuel and wear and tear on generator sets, since they run only when needed.

Ease of expansion is another reason for paralleling. If the loads in a distribution system are expected to expand substantially, the initial investment can be minimized by installing one smaller generator set, then adding more sets in parallel as loads increase. Often, emergency power capacity can be added with less disruption with a paralleling application.


For standby systems, the most important reason for paralleling is increased reliability of service to critical loads. When there’s a normal power outage, all generators in a system are started. The first generator set ready to handle the critical load does so. As the other sets approach nominal voltage and frequency, they are connected to the bus and additional loads can be connected in priority order. If a generator fails, less critical loads are shed, but the most critical loads still get emergency power.

A related benefit of paralleled generator sets is servicing convenience. When a generator is taken out of service for maintenance or repair, other generators in the paralleled system are still available to provide power. Maintenance and service for the generator sets can be provided without taking the entire generator system out of service.

Load Sharing

Another advantage of installing generator sets capable of being paralleled is that there may be economic advantages to operating the generator set in parallel with the local utility, reducing facility demand. Many utilities have distributed generation programs that provide economic incentives for large power users to decrease electrical demand at certain times of the day. For example, a large industrial facility with standby generator sets with paralleling capabilities would be able to significantly reduce their electrical load on the utility over the peak demand periods. By running their generator sets in parallel with the utility, they would be able to negotiate a special rate, thereby reaping savings on their power consumption.

Each utility has protocols and rates for self-generated power and paralleling to reduce peak demand. Operationally, the utilities have established standards for interconnection that deal with voltage, frequency control, power quality, safety and metering. End users interested in paralleling for load sharing and cost saving should consult their local utility for more information.

Parallel Considerations

Paralleling is not the best solution for every generating system. Each case must be examined individually to determine the best set-up for the generating system to be used.

When to parallel – Paralleling should be considered when there is a need for increased reliability of service to critical loads. It should also be considered when total emergency power needs exceed the capacity of a single generator set. Assuming there are needs for high reliability and multiple generator sets for increased capacity, then local utility load sharing programs offer significant economic benefits.

If an end user has different kinds of loads, but only a portion of them are true emergency loads, paralleling is an economical option. In emergency situations, the backup power system is put into service and one generator set will be able to handle the initial critical load as other generator sets are put into service to handle the non-critical loads. In some situations, the non-critical load can just shut down and the critical loads will be the only ones to draw from the power source. A paralleled system makes either of these options viable.

In the same situation, if all of the end user’s total load requirement is greater than the capacity of one generator set, the end user may parallel two or more generators sets, operating at lower power, to handle the critical and non-critical loads during an emergency outage. A single set would not have the capacity to handle the entire load, but working together, they could handle the load with some power to spare.

When Not to Parallel – In some cases, the added reliability and flexibility provided by paralleled generator sets are not important enough to be worth the additional costs required to set up a parallel system. If temporary power outages do not result in serious problems, the loads on the distribution system may not be critical enough to demand the extra reliability of a paralleled system.

One big difference between a paralleling system and a single large generator set is that with a paralleling system, load carrying capacity increases as generator sets close to the bus, while with a single genset system, usually all the capacity is immediately available when the genset starts. Because of this, for paralleling applications the facility loads are usually split into load blocks based on the priority of service to the specific load. For example, the emergency loads (such as egress lighting) are priority one, important mechanical loads would be priority two, and “nice to have” loads would be priority three. When the power fails, the system is designed so that the first generator set available serves the emergency loads, and less critical loads must wait for service until more generator capacity is available. A consequence of this is that first-priority loads are usually served more quickly and reliably than low-priority loads. The generator set size is selected so that any one generator set can serve all the emergency (priority one) loads.

In systems where the facility loads can’t be conveniently split or controlled by their priority, the advantage of a paralleling system may be lost. For example, in a data center, for the facility to be operational, both the UPS equipment and air conditioning equipment must be operational. If it takes two generator sets to run these loads, and one genset fails, the whole system would eventually fail. In order to retain improvements in reliability in systems of this type, a redundant generator set is often provided (three gensets are provided, rather than two).

If an application has existing generator sets and those generator sets are not compatible for paralleling, then one or more generator sets must be replaced with compatible ones in order to implement a paralleling system. The cost of this replacement may be more than the value of the increased reliability or other advantages of a paralleled system. p

Guidelines for Paralleling

Prime Movers (Engines) – Identical engine models should be used for new paralleling installations, because engines with differing characteristics may result in nuisance failures.

Generators – Generators used for paralleling must have compatible voltage waveforms. When generators have the same pitch (normally 2/3), the difference in the harmonics caused by voltage waveform differences should not cause problems.

Governors and Voltage Regulators – Governors and voltage regulators should be identical on all the generator sets to avoid load sharing problems in the system.

Load Control – When two or more generator sets are paralleled, each generator set should provide its proportionate share of the power to the load. The amount of real load served by a generator operating in parallel with other units is a function of engine output, whereas the portion of the reactive load supplied is determined by the internal driving voltage (excitation) of the individual power source. In a paralleling situation, the governor controls the real (kW) output of the generator. The voltage regulator controls the flow of reactive power (kVAR). Paralleling applications require governor and voltage regulation controls that work together. In addition, the system must manage the sequence of load adding and load shedding in the facility power distribution system. This load management plan is the single biggest difference between single and parallel generator applications.

Synchronizer – A synchronizer is essential for paralleled systems to monitor the phase relationship between two voltage sources and provide a correction signal to an engine governor to force the generator set to synchronize with the system bus.

System Protection – System protection for generator systems consists of devices to help protect the system from possible damage during operation and to retain the integrity and reliability of the system. These systems and their ANSI standards designations include:

• Generator circuit breakers, ANSI Device 52

• Loss of Field/Reverse VAR, ANSI Device 40

• Reverse power protection, ANSI Device 32

• Permissive paralleling protection, ANSI Device 25

• Paralleling suppressors, which protect the excitation system from voltage surges that can occur when paralleling.


Gary L. Olson is Technical Counsel for Cummins Onan and is the acting product manager for paralleling and generator set control products. His work experience covers a wide range of sales, marketing and application engineering positions. Employed by Cummins Onan since 1977, he holds a BS degree in mechanical engineering from Iowa State University and an MBA from the University of St. Thomas in St. Paul.