# Evaluating Harmonic Concerns With Distributed Loads

When engineers need to evaluate harmonic concerns on power systems, they generally refer to Standard 519-1992 from the Institute of Electrical and Electronics Engineers (IEEE). But in distributed-load situations, questions frequently arise while trying to apply the recommended limits. That's what happened when Bob Worden, an engineer with New York State Electric and Gas (NYSEG), tried to apply the limits to a ski resort. His questions illustrate some of the general concerns for employing harmonic limits with distributed sources.

When engineers need to evaluate harmonic concerns on power systems, they generally refer to Standard 519-1992 from the Institute of Electrical and Electronics Engineers (IEEE). But in distributed-load situations, questions frequently arise while trying to apply the recommended limits. That's what happened when Bob Worden, an engineer with New York State Electric and Gas (NYSEG), tried to apply the limits to a ski resort. His questions illustrate some of the general concerns for employing harmonic limits with distributed sources.

Worden submitted a question to the PQ Group (www.pqagroup.com). He wondered what he should use as the rated load current (Il) in applying Standard 519-1992's limits to a ski resort. Il is important for two reasons: It's used to determine the short-circuit ratio at the point of common coupling, and it's the base current for applying the harmonic limits.

Standard 519-1992 says to take the maximum demand current each month for 12 months and average those values. But a ski resort might operate for only five to six months out of the year. So should Worden use the demand currents from the operating months alone or the whole year?

Here's a few more questions associated with evaluating harmonic limits at ski resorts (or for that matter, any system with distributed loads). Where is the point of common coupling? Should the harmonic limits be applied at the supply to each individual lift? Typically, separate locations from the distribution system supply power to each lift. But should the ski resort be evaluated as a single facility? This is a critical distinction because if you consider the resort as a whole, there will be harmonic cancellation between the different lifts.

Of course, Worden's ultimate objective is to make sure the system voltage distortion does not cause problems at the ski resort or other contiguous locations.

Several years ago, Electrotek engineers evaluated harmonic issues at a ski resort. The remainder of the article uses this past evaluation to highlight the different options for controlling harmonic distortion levels at a ski resort and the advantages and disadvantages of each approach. As you read about this case study, think about applying the knowledge gained to other distributed-load situations.

### Harmonic Sources

Ski lift drives are the most important sources of harmonics at ski resorts. These drives are usually variable-speed and either DC or AC. In both cases, they are relatively large loads with significant harmonic generation. Fig. 1a shows a typical current waveform from a ski lift's DC, 6-pulse drive — the most common configuration for individual lift drives. Fig. 1b illustrates the same current's harmonic spectrum.

AC drives are likely to have similar harmonic generation characteristics. The main difference is that the AC drives probably employ a diode bridge for rectifying the AC voltage to DC. DC drives perform this rectification with silicon-controlled rectifiers (SCRs). The SCRs are phase-controlled, which can lead to some cancellation of harmonic components (injected by multiple drives around the system). The level of cancellation is lower with the AC drives because all the drives have diode bridges (without control) for the rectification.

Another important difference between the two types of drives involves the displacement power factor (DPF). With the phase-controlled operation of the SCRs, the DC drives can exhibit low power factor. This increases the losses in the step-down transformer and can result in voltage regulation problems at the drive location. The low power factor and increased losses may justify the use of harmonic filters at the drive location. AC drives will have a much better power factor, making the use of harmonic filters at the drive location less attractive.

Another important consideration is how the harmonics from the different drives combine on the system and result in voltage distortion. Ski resorts are usually supplied from relatively long distribution circuits, which can increase the concern for high voltage-distortion levels. Ski lifts will be spread out across one or more distribution feeder circuits. The concern for voltage distortion levels is the result of the harmonic currents injected by all the lifts in combination with the distribution system's frequency response characteristics.

Fig. 2, on page 10, illustrates the supplies to ski lift loads spread out across a distribution system. The various lift sizes are indicated. The harmonic currents from these drives combine on the distribution system and flow through the system impedance, causing voltage distortion. This can be a particular problem if the distribution system exhibits resonance conditions that magnify the harmonic currents.

### Voltage Distortion

Two possible elements can lead to high voltage-distortion levels. The first is a relatively weak power system. Ski resorts are found in remote locations and supplied through long feeder circuits, which increases system impedance. The harmonic voltage distortion is a function of this impedance.

The second, and more common, element correlates with resonance conditions on the distribution system. Resonances could be caused by power factor correction capacitors used to control the voltage and reactive power flows on the system, or they could be associated with the system itself. Ski resorts often have significant lengths of underground cable, and the capacitance associated with the cable results in resonance conditions.

In Fig. 3, you'll see the frequency response characteristics for such a system. The utility has not added power factor correction capacitors. The resonance comes from the cable capacitance of the feeder circuits in combination with the source inductance. You can obtain this kind of information by simulating a system response using a software tool like SuperHarm. You also can use the tool to try out different options for changing a system's frequency response or reducing the harmonic injection.

Fig. 3 also shows a resonance that can cause magnification in the range of the 13th to 19th harmonics. These are not the dominant harmonics for 6-pulse DC drives, but the resonance can make the voltage distortion at these frequencies excessive. In the figure, voltage distortion levels exceed IEEE 519 guidelines, or 5% total harmonic distortion. The highest voltage distortion levels occur at the individual drive locations (see Fig. 4), but the voltage distortion levels exceed recommended levels in the whole resort village. One common complaint is that UPS systems incorrectly interpret the distorted voltage as interruptions. They then switch to batteries until the batteries completely discharge. The high voltage-distortion levels also produce heating in motors and transformers, and they can cause some electronic equipment to misoperate.

### Controlling Harmonic Distortion

What's the best way to control harmonic distortion levels on systems that supply ski resorts? I've summarized three options here, but the best could be any one of these, depending on the specific circumstances.

Applying harmonic filters at individual lifts. With DC drives, filters provide power factor correction and harmonic control at the 480V bus, reducing the harmonic currents injected into the power system. This is an attractive option for DC-drive loads because of the harmonic control, plus reduced losses and improved voltage control at the low-voltage bus. It's usually sufficient to apply a single-tuned filter (usually tuned to slightly below the 5th harmonic) at each large lift drive. This improves the harmonic performance of the entire system. As a rule of thumb, lifts larger than 500 hp should have individual filters.

Applying harmonic filters on the distribution system. When there are distributed harmonic sources on the distribution system, it may be more economical to control the harmonic levels with filtering on the actual distribution system.

I recently worked with Carolina Power & Light to design a substation filter that combines the power factor correction required at the substation with harmonic control. This filter solved a resonance problem on the distribution system that produced voltage-distortion levels exceeding 7% — despite the fact that individual customers on the system were typically not exceeding IEEE 519 harmonic current injection limits.

Ski resorts are prime examples of distributed harmonic sources on the distribution system. The lift drives are spread out over a wide area, making it possible to control the harmonics with strategically placed filters on the distribution system. This usually requires a study to determine the best place for the filters, the size of the filters, and appropriate tuning.

Fig. 5, on page 12, shows a filter configuration that includes a 5th harmonic and a 7th harmonic tuned branch. The effect on the system frequency response is shown in Fig. 6. Note that the response still exhibits a higher frequency resonance, which could be a concern. You could resolve this problem by adding a shunt capacitor branch in parallel to the tuned filters.

Using cancellation to reduce the harmonic levels on the system. The final option involves designing the system so that harmonic current components from different lift locations cancel each other out on the distribution system. If some of the lift drives are supplied with delta-delta transformers (or wye-wye transformers) and some are supplied with delta-wye transformers, there will be natural cancellation of 5th and 7th harmonic components due to the phase shift introduced. This cancellation is not ideal in the case of DC drives used for ski lifts. That's because DC drives operate with different characteristics and are separated by lengths of distribution circuit, which introduces additional impedance and phase shift. However, you can reduce the harmonic distortion levels on the overall system to acceptable levels.

This technique is particularly appropriate when the lifts use AC drives because the diode bridges do not introduce additional phase shift like the controlled rectifiers do. This final option was the technique employed in our evaluation several years ago, and it reduced the overall voltage distortion to acceptable levels on the entire distribution system.

### Conclusion

Ski resorts illustrate important questions about how to apply harmonic limits according to IEEE standards and design effective solutions for systems with distributed loads. The distributed nature of the harmonic sources requires close cooperation between resort operators and distribution system companies so voltage-distortion levels are kept to acceptable levels on the overall system. Working together, it is often possible to design a more economical solution. This may require a flexible interpretation of the harmonic limits in IEEE 519-1992. Every system is unique, and cooperation is the best way to make sure you select the optimum solution.

Members of the IEEE 519A task force are attempting to emphasize such flexibility in a new guide. The purpose of the document is to set guidelines for applying harmonic limits in a variety of applications. The IEEE is currently finalizing this new guide through its balloting process.

Mark McGranaghan directs power quality projects and product development at Electrotek Concepts. He also works on the IEEE 519A task force. Contact him at [email protected].