Short T.A. Electric Power Distribution Handbook

Подождите немного. Документ загружается.

Distributed Generation

Distributed generation (DG) or embedded generation (the European term)

refers to generation applied at the distribution level. EPRI (1998) deﬁnes

distributed generation as the “utilization of small (0 to 5 MW), modular

power generation technologies dispersed throughout a utility’s distribution

system in order to reduce T&D loading or load growth and thereby defer

the upgrade of T&D facilities, reduce system losses, improve power quality,

and reliability.”

No exact size or voltages are accepted as deﬁnitions of distributed gener-

ation. A draft IEEE standard (IEEE P1547-D11-2003) applies to generation

under 10 MW. Distribution substation generation is normally considered as

distributed generation; sometimes subtransmission-level generation is also

considered as DG since many of the application issues are the same. Related

terms — non-utility generator (NUG) and independent power producer

(IPP) — refer to independent generation that may or may not be at the

distribution level. A broader term, distributed resources (DR), encompasses

distributed generation, backup generation, energy storage, and demand side

management (DSM) technologies.

Smaller-sized generators continue to improve in cost and efﬁciency, mov-

ing closer and closer to the performance of large power plants. At the same

time, utilities face signiﬁcant obstacles when building large facilities — both

power plants and transmission lines. Utilities or end users can install mod-

ular distributed generation quickly. This local generation reduces the need

for large-scale utility projects. Distributed generation can allow utilities to

defer transmission and distribution upgrades. Also, DG reduces losses and

improves voltage. With the right conﬁguration, distributed generation can

also improve customer reliability and power quality.

Both utilities and end users can install distributed generators, usually for

different reasons. Distributed generators are most cost effective if the cus-

tomer has need for

•

Cogeneration

— Using the generator waste heat locally signiﬁcantly

improves the economics of many applications. Uses include process

steam, heating water, facility heat, or running air conditioning.

1791_book.fm Page 699 Monday, August 4, 2003 3:20 PM

700

Electric Power Distribution Handbook

•

Reliability

— Many locations need backup generation for reliability

purposes. Once they are bought, running them full time or for peak

shaving becomes more economical.

Utilities most need local generation with

•

Overloaded circuits

— Distributed generation can delay infrastructure

upgrades.

•

Generation shortages

— If a utility needs generation anywhere it can

get it, the utility or end users can quickly install distributed generation.

Applying generation closer to the load beneﬁts the transmission and dis-

tribution infrastructure. Local generation can relieve overburdened trans-

mission and distribution facilities as well as reduce losses and voltage drop.

While offering beneﬁts and opportunities, generation is not always easy

to integrate into existing distribution systems (Barker and De Mello, 2000;

CEA 128 D 767, 1994; EPRI 1000419, 2000). Distribution systems were never

designed to include generation; they were designed for one-way power ﬂow,

from the utility substation to the end users. Generators violate this basic

assumption, and generators can disrupt distribution operations if they are

not carefully applied. One of the most critical situations is that a distribution

interrupter may isolate a section of circuit, and the generator might continue

supplying the load on that section in an

island

. Islanding poses safety haz-

ards, and islanded generators can cause overvoltages on the circuit. In addi-

tion to islanding, generators can disturb protection, upset voltage regulation,

and cause other power quality problems. We will investigate many of these

issues in this chapter, how to identify problems and options for ﬁxing them.

In addition to the technical difﬁculties, distributed generation raises sev-

eral other issues. How do we meter a generator? What is a fair price to pay

a generator injecting power into the system? How can distributed generators

be dispatched or controlled, especially if they are owned by end users? How

can we apply the generators to optimal locations on the system, rather than

just accepting it wherever end users install it?

Many distributed generation technologies can supply isolated loads on

their own, but massive disconnections from the utility to stand-alone systems

are unlikely in the near term. The utility distribution system solves many of

the problems with stand-alone generation systems:

•

Reliability

— Generators must be applied in redundant fashion to

achieve the reliability of normal distribution service. Together, a

utility service along with local generators has much higher than

normal reliability.

•

Matching load —

Many small generators are most economical when

they run at full power all of the time, but most loads have an average

power demand that is well below the peak. A utility source allows

the generator to run at full power, and the utility can offset the power

1791_book.fm Page 700 Monday, August 4, 2003 3:20 PM

Distributed Generation

701

for the load. In addition, much of the renewable generation is inter-

mittent. Without a utility interconnection, renewable sites must have

energy storage, which is expensive.

•

Load following

— Many distributed generation designs have limited

load-following capability. Step changes in load or motor starting can

deeply affect the voltage and frequency.

The utility supply has the beneﬁt of tying thousands of generators together,

providing reliability and matching load. In the future, small generators may

be applied locally in ways that gain some of these same beneﬁts as the bulk

utility system.

Microgrids

tie generators together on a localized scale to gain

many of the reliability and load matching advantages of the traditional utility

grid (EPRI 1003973, 2001).

Distributed generation technologies continue to advance: cost comes

down, performance improves. Projections of penetration of distributed gen-

eration into the electrical system vary widely. Since natural gas delivers

energy at a cost that is roughly one fourth of the cost of electric energy, if

an efﬁcient, low-capital-cost, and low-maintenance distributed generator

becomes available, gas energy delivery has an advantage over electric energy

delivery. Utilities must prepare for several scenarios and consider distributed

generators as another tool for supplying end users with electric power.

14.1 Characteristics of Distributed Generators

14.1.1 Energy Sources

Several energy sources drive distributed generators including

• Reciprocating engines

• Combustion turbines

• Microturbines

•Wind turbines

• Fuel cells

• Photovoltaics

Three power converters convert the power output of the energy source to

interface with standard 50- or 60-Hz systems:

• Synchronous generator

• Induction generator

• Inverter

1791_book.fm Page 701 Monday, August 4, 2003 3:20 PM

702

Electric Power Distribution Handbook

Normally, reciprocating engines and combustion turbines interface

through synchronous generators. Microturbines, many wind turbines, fuel

cells, and photovoltaics interface through inverters.

Three factors are considerations for the energy source: the stability of the

power source, the ability to follow changes in load, and the reliability (avail-

ability) of the source.

On reciprocating engines, combustion turbines, and microturbines, the

burn rate of the fuel/air mixture controls the power output. The real power

output is controllable and responds to load changes fairly quickly. These

combustion-based generators are normally very stable sources (unless the

fuel supply is poor).

Wind turbines and photovoltaics produce variable output that is nor-

mally not controllable. One of the main concerns of wind turbines on

distribution circuits is the possibility of voltage ﬂicker due to pulsating

output, mainly because the tower shades the blades for part of their rota-

tion. Sunshine variability also causes variation in solar-panel outputs, but

the changes are normally slow enough to limit ﬂicker problems. Neither

of these can follow load.

Fuel cells have load-following capability, but it is signiﬁcantly slower than

combustion generators. Fuel cells and photovoltaics have no natural provi-

sion to provide short-time overloads (some manufacturers include batteries

or ultracapacitors to add this capability). This helps control fault currents

but limits their ability to supply motor starting current and respond to loads.

The availability and reliability of the energy source plays a role in some

applications. Combustion engines and turbines are easily dispatchable. Pho-

tovoltaics and wind generation are non-dispatchable; the energy source is

not always available. Small hydro units may or may not be dispatchable,

depending on the water levels.

14.1.2 Synchronous Generators

The workhorse power generator for over a century, the synchronous gener-

ator is the power converter of choice for most distributed power sources.

Reciprocating engines and gas turbines almost always interface with a syn-

chronous generator. A synchronous generator can operate independently or

in synchronization with the utility. Most distribution-scale generators are

salient-pole machines with four or six poles driven by 1800- or 1200-rpm

reciprocating engines (with some even slower in larger units). Distribution-

scale gas-turbines spin faster but are normally geared down to drive a salient-

pole generator at 1800 rpm.

Synchronous generators can operate at leading or lagging power factor and

can operate in a voltage control or voltage following mode. Most distributed

synchronous generators use a voltage-following control mode. This differs

from large generators on the electric system, which have controls to regulate

voltage (within the limits of the var capability). Distributed generators nor-

mally follow the utility voltage and inject a constant amount of real and reactive

1791_book.fm Page 702 Monday, August 4, 2003 3:20 PM

Distributed Generation

703

power (so the power factor stays constant). Normally, most DG operators run

at unity power factor since that produces the most watts for a given kVA rating.

The three-phase ac power system connects to the armature winding on

the stator (the outside, nonmoving part) of the synchronous machine (Figure

14.1). The stator wraps around the outside of the generator, so it can couple

magnetically with the rotating magnetic ﬁeld produced on the rotor. The

rotor has a dc magnetic ﬁeld rotating at synchronous speed. The rotor has

several poles around the shaft. Each pole is a piece of iron with a coil

wrapped around it (the ﬁeld winding). The ﬁeld winding forms an electro-

magnet that creates the dc magnetic ﬁeld. A separate device, an exciter,

generates the dc current for the ﬁeld winding. The rotating dc magnetic ﬁeld

induces an emf in the stator (armature) winding, which drives power into

FIGURE 14.1

Synchronous generator. (Courtesy of Caterpillar, Inc.)

1791_book.fm Page 703 Monday, August 4, 2003 3:20 PM

704

Electric Power Distribution Handbook

the ac system. For more detailed descriptions, see one of the many machine

books available, such as Fitzgerald et al. (1971).

A voltage source (the emf induced on the armature) behind a reactance

models a synchronous machine well. Several impedances come into play

depending on the scenario:

•

Subtransient reactance

—

" is the positive-sequence impedance that

determines the three-phase fault current during the ﬁrst few cycles.

Normally,

" is between 10 and 30% on the machine rating. Higher

values of

" limit the short-circuit current the machine supplies but

reduce the capability of the machine to respond quickly to load

changes.

•

Negative-sequence reactance

—

impacts the current under voltage

unbalance.

is equal to the average of

" and

" (the direct and

quadrature impedances) and is typically 10 to 30% on the machine

rating.

•

Zero-sequence reactance

—

affects ground-fault currents and the

ﬂow of triplen harmonics. The zero-sequence reactance is normally

very low, even less than 5% on the machine rating.

In the absence of better information, assume

" = 20%,

= 25%, and

= 5% on the machine rating. In the past, this information was hard to obtain

for small machines, but now most generator manufacturers supply them.

Table 14.1 shows values of small generator constants from tests, with the

characteristics of machines C and D being most representative. Most of the

parameters in the table are only needed for very detailed simulations.

For fault calculations, the voltage behind the subtransient reactance is

normally in the range of 1.0 to 1.3 per unit with a typical value of 1.1. The

power factor ranges from 0.8 lagging to unity. So, with the voltage at 1.1 per

unit and

" = 20%, the initial three-phase fault current is 5.5 times the

generator rating.

Under a fault, sudden load change, or other transient condition, the sub-

transient reactance (

" which is often denoted as just

") dominates. After

a time equal to the subtransient time constant (

"), which ranges from 1 to

10 cycles, the transient reactance (

) more accurately models the generator

response.

The direct-axis impedance (indicated by the subscript

) is the effective

impedance for changes that are in sync with rotation (50- or 60-Hz events).

Quadrature-axis impedances (subscript

) are the impedances seen with the

rotor out of phase with the stator ﬁeld. For fault calculations we only need

to consider the direct-axis impedances. For unbalance or harmonics, the

quadrature-axis impedances have impacts.

The

saturated

subtransient reactance (

") is used for short-circuit studies

and for harmonics. The subscripted “

” means constant voltage; the imped-

ance is the value of

V/I

at rated voltage and no load for a short circuit on

1791_book.fm Page 704 Monday, August 4, 2003 3:20 PM

Distributed Generation

705

the generator terminals (which is a high-current condition that saturates the

magnetic core, which results in a lower impedance). Another version of

subtransient reactance, the unsaturated reactance (

") is used for stability

studies. Usually, if it is not indicated which value it is, it is the saturated

value,

The negative-sequence reactance (

) and the subtransient reactance (

are determined mainly by the damper windings on the rotor. On a salient-

pole machine, damper windings are a shorted winding wound around each

pole. Damper windings react to dampen out transients on the machine. The

torque produced by the damper windings slows down a machine that is

running faster than synchronous speed, and it speeds up a machine that is

running slower than synchronous speed. Damper windings are like the rotor

on a squirrel-cage induction machine (in fact a squirrel-cage induction

machine is a synchronous machine with damper windings and no ﬁeld on

the rotor). At synchronous speed, no current ﬂows in the damper windings

since the ﬂux on the rotor is dc, but if transients act to change the frequency

relative to the synchronous speed, currents ﬂow in the damper windings

and create torque to counteract the disturbance. The damper windings also

TABLE 14.1

Small Synchronous Generator Constants in per Unit

on the Machine Base

Machine A B C D

kVA 69 156 781 1044

kW 55 125 625 835

V 240/480 240/480 240/480 240/480

rpm 1800 1800 1800 1800

gen

0.26 0.20 0.40 0.43

pf 0.80 0.80 0.80 0.80

2.02 6.16 2.43 2.38

0.171 0.347 0.254 0.264

0.087 0.291 0.207 0.201

1.06 2.49 1.12 1.10

0.163 0.503 0.351 0.376

(sec) 0.950 1.87 1.90 2.47

(sec) 0.080 0.105 0.198 0.273

(sec) 0.078 0.013 0.024 0.018

(sec) 0.004 0.011 0.020 0.014

(sec) 0.045 0.020 0.016 0.009

(sec) 0.007 0.004 0.005 0.003

0.011 0.034 0.017 0.013

(sec) 0.014 0.022 0.038 0.032

0.038 0.054 0.051 0.074

0.125 0.375 0.279 0.260

Source:

Gish, W. B., “Small Induction Generator and Syn-

chronous Generator Constants for DSG Isolation Studies,”

IEEE Transactions on Power Systems

, vol. PWRD-1, no. 2,

1791_book.fm Page 705 Monday, August 4, 2003 3:20 PM

706

Electric Power Distribution Handbook

react to other stator currents that are not 50- or 60-Hz positive-sequence

currents (signiﬁcantly, negative-sequence currents and harmonic currents).

The negative-sequence impedance determines the current ﬂow due to

negative-sequence voltages. The impedance to harmonic voltages is also

approximately the negative-sequence impedance. The negative-sequence

impedance is the same whether the generator is under steady-state, transient,

or subtransient conditions.

The zero-sequence impedance of a synchronous machine can have

extremely low impedance. It is enough of a problem that many generators

are ungrounded or grounded through an impedance to prevent the ﬂow of

zero-sequence current. Many generators are not braced to handle the fault

current for a line-to-ground fault at the terminals of the machine. Single-

phase faults cause more mechanical stress and are higher magnitude. Ground

fault currents are 30 to 40% higher than three-phase fault currents (

" vs.

/(2

)

1.3 to 1.4

"). The zero-sequence impedance is the same

whether it is under steady-state, transient, or subtransient conditions.

The reason that the zero-sequence impedance is so low is that magnetic

ﬁelds from zero-sequence currents in the stator winding tend to cancel each

other. If the ﬁelds cancel and couple very little to the rotor, the impedance

is very low.

The zero-sequence impedance varies signiﬁcantly with design. The most

prominent difference is due to the pitch of the stator winding. A pole pitch

is the number of degrees that the rotor has to move to change from one pole

to the other. In a 2-pole machine, one pole pitch is 180

∞

, and in a 4-pole

machine, it is 90

∞

. The

pitch factor

(or just the

pitch

) of the stator winding is

the portion of the pole pitch that the stator winding spans. A full-pitch stator

winding spans the full pitch. A fractional pitch winding spans less than the

full pitch. Figure 14.2 shows a comparison of a full-pitch winding and a 2/

3-pitch winding. In this ﬁgure, each phase has two windings (

and

for

example); the current in each winding goes out on one slot conductor (like

) and returns in another slot conductor labeled with the prime notation

(like

). The 2/3-pitch winding reduces the zero-sequence impedance the

most. Because the two conductors in each slot have current in opposite

directions, the ﬁelds cancel almost completely (since

a = b = c =

for zero-sequence current). Other common pitch factors are 5/8 and 3/4.

Synchronization is important when connecting synchronous generators. If

a large machine is out of synch with the utility when it is connected, the

sudden torque can damage the generator, and the connecting switch and

other devices will have a large voltage disturbance. To avoid these problems,

synchronizing relays are required for synchronous generators to ensure that

the voltage, frequency, phase angle, and phasing are the same on the utility

and generator. Normally, the generator is brought up to speed, and the ﬁeld

current is adjusted to bring the voltage close to the utility system. The

frequency is more precisely adjusted to bring it within 0.5 Hz of the power

system. Then, the synchronizing relay allows closing when the two voltages

are within 10

∞

(CEA 128 D 767, 1994).

1791_book.fm Page 706 Monday, August 4, 2003 3:20 PM

Distributed Generation

707

14.1.3 Induction Generators

Induction generators are used in a few distributed generation applications,

mainly wind-turbine applications where the speed of the prime mover varies.

Induction generators are simpler than synchronous generators. They do not

have exciters, voltage regulators, governors, or synchronizing equipment.

An induction generator is the same as an induction motor except the prime

mover turns the rotor faster than synchronous speed. An induction generator

needs supplemental excitation, either supplied by the utility system or local

capacitors.

To start some induction generators, the utility may accelerate the prime

mover to operational speed, drawing inrush to ﬁve or six times the generator

rating just like motor starting. A reduced-voltage starter using a reactor or

autotransformer reduces the inrush. If possible, the prime mover can accel-

erate the shaft to near synchronous speed before closing in; this softens the

inrush when connecting. This type of connection draws inrush of up to three

times the generator rating. Because the induction machine cannot generate

voltage on its own, the timing of the connection is unimportant.

Even though the induction generator cannot create voltage on its own, the

generator can self-excite with capacitors. Reclosing into a self-excited gen-

erator with local capacitors can damage the shaft and other equipment.

Under transient conditions, induction machines respond similarly to syn-

chronous machines. Under fault-type transient conditions,

= 20% (the

locked-rotor impedance) and

= 5%. The voltage behind the positive-

sequence impedance is also about 1.1 per unit. The main difference between

synchronous and induction machines is that the generator driving voltage

collapses more quickly in an induction machine.

FIGURE 14.2

Comparison of a full-pitch and a 2/3-pitch stator winding for a two-pole machine with a double-

layer stator winding with one slot per phase per pole.

Full-pitch winding Two-thirds-pitchwinding



1791_book.fm Page 707 Monday, August 4, 2003 3:20 PM

708

Electric Power Distribution Handbook

14.1.4 Inverters

Many types of small generators best interface with utility and utilization

power through an inverter. Photovoltaics and fuel cells generate dc. Wind

turbines and microturbines generate incompatible frequencies, so they are

best applied by rectifying to dc then converting back to ac with an inverter.

Two main classes of inverters are

• Line commutated

• Self commutated

A three-phase line-commutated inverter is the simplest and least expensive

converter. Thyristors (also called silicon-controlled rectiﬁers or SCRs) in a

full-wave bridge conﬁguration convert from dc to ac by switching current

(see Figure 14.3). Thyristors can switch large amounts of current at low loss.

The inverter controls when to turn on the thyristor by applying a voltage

pulse to the gate of the thyristor; the thyristor does not turn off until the

next current zero.

Since a line-commutated inverter draws currents in square waves, it creates

large harmonics, primarily the 5th, 7th, 11th, and 13th. Line-commutated

inverters require signiﬁcant reactive power from the system, between 10 and

40% of the inverter rating. It is possible to use 12-, 18-, or 24-pulse bridge

designs with phase-shifting transformers to reduce harmonics.

Line-commutated inverters do not provide signiﬁcant fault current. The

reactor in the inverter limits the fault current. Most designs limit current to

100 to 400% of the inverter rating. Usually inverters have overcurrent pro-

tection that removes the ﬁring pulses when current exceeds some threshold.

The duration of fault current output also depends on the amount of capac-

itance on the inverter dc bus. Note that a harmonic ﬁlter on the output can

increase the momentary fault value.

The self-commutated inverter can generate its own sinusoidal voltage.

Although several self-commutated conﬁgurations are around, the most com-

mon is the fast switching pulse-width modulation (PWM) design using

insulated gate bipolar transistors (IGBTs) or other switchable power elec-

tronic components. IGBTs can be turned on and off quickly. By controlling

FIGURE 14.3

Common three-phase bridge inverter line-commutated conﬁguration with thyristors.

dc source

1791_book.fm Page 708 Monday, August 4, 2003 3:20 PM