Documente Academic
Documente Profesional
Documente Cultură
Lecture Notes # 11
Prof. S. S. Venkata
Load Modeling
Under steady-state, complex power S at any location in a distribution
system is a function of voltage, V
S = VI* = P + jQ = f ( V )
Model I
It is usual to represent both P & Q as a polynomial function of |V|. That
is
P = ao + a1|V| + a2|V|2 + a-1|V|-1 +
(1)
Q = bo + b1|V| +b2|V|2 +
(2)
Q = b0
CONSTANT POWER
MODEL
Q = b1 | V|
2
Q = b 2 | V|
Example [1]
(a) Air-conditioning load demand (per unit values)
P = 2.97
2
4.00 |V| + 2.02 |V|
Q = 12.90
2
15.60 |V| + 10.3 |V|
Model II
Composite loads which are assumed to be mixtures of the types
discussed above can be represented as
| V|
P = Pn
| Vn |
| V|
Q = Qn
| Vn |
, k varies between 0 to 3
(3)
, l varies between 0 to 3
(4)
Examples [2]
(a) If k = 1, l = 0, it implies the load is a constant current type, with
power factor unity.
(b) If k = l = 2, then the load is constant impedance type.
(c) If k = 2.5 and k = 2.7, then it represents an aluminum reduction
plant. This model is simpler and can be determined empirically knowing
|Vn| = initial or base value of voltage
Pn = initial or base value of real power
Qn = initial or base value of reactive power
Note: In all these models, one should be aware of the range of |V| for
which the equations (1) to (4) are valid.
Model III
Loads are seldom modeled to include frequency effects. If loads (or
demands) are sensitive to frequency effects, then these effects should be
included: In this case
| V|
P( f ,| V| ) = Pn
n | Vn |
| V|
Q( f ,| V| ) = Q n
n | Vn |
where,
(5)
= 2( f )
n = 2( f n )
f n = base frequency
(6)
, = constant exponents.
With this type of model, it is often a practice to determine four sensitivity
coefficients defined below:
dP
df
dP
d| V|
dQ
df
, and
dQ
d| V|
(dP / df ) (dp / d| V| ) f
( dQ / df ) ( dQ / d| V| ) | V|
Then: P = Pn + P
Q = Qn + Q
for
f = fn + f
and
The following tables show the results of a data survey conducted by EPRI
[3].
Model IV [4]
This is particularly suitable for modeling aggregate loads at a node
knowing the demand profiles for a day, a season or a year.
Assuming Gaussian Distribution
Pp = Pn + k p
(7)
Q p = Q n + k p
(8)
References
[1] R. B. Adler, C. C. Mosher,
Power Loads
, Vol. I,
Demand:
Diversity factor, FDV is the ratio of the sum of individual maximum demands of
various subdivisions of a system to the maximum demand of the whole system. It is
greater than or equal to 1. This means it is the ratio of maximum non-coincident demand
to maximum coincident demand.
Coincidence factor, FC is the ratio of the maximum coincident total demand of a
group of consumers to the sum of the maximum power demands of individual consumers
comprising the group both taken at the same point of supply for the same times.
Coincidence factor is the reciprocal of the diversity factor.
Load diversity, LD is the difference between the sum of the peaks of two or more
individual loads and the peak of the combined load.
(1)
D m1 + D m 2 + + D mN =
D mi = D m
Dm
Fc
i =1
D m1 + D m 2 + + D mN =
D mi = Dm
Dm
Fc
i =1
(22)
where N = Number of loads connected
1
1
FDV FDV
(2)
Coincidence factor = FC =
(3)
Load diversity = LD = DM - Dm
(23)
(24)
The maximum demand of the group of (1+2+3+. . . . +N) loads can also be expressed as:
Dm = c1D1 + c2D2 + c3D3 + . . . . . . .+ cNDN
(25)
where cN =
the contributing factor of the Nth load which is defined as the contribution
of the nth load to the group maximum demand. It is equivalent to
hourly
listed in Table 2.
variation factor
(26)
D1 + D2 +L L + DN
If D1 = D2 = . . . . . . = DN then
Fc =
D1 ( c1 + c2 +L + c N )
ND1
c1 + c2 +L + c N
N
(27)
(28)
If c1 = c2 = . . . . . . = cN then
Fc =
c1 ( D1 + D2 +L + DN )
= c1
D1 + D2 +L + DN
Loss Factor:
Ratio of average power loss to peak load power loss during a specified
period of time.
Referring to Figure 3
x = peak load of duration t
y = minimum load of duration (T - t)
Average load =
xt + y (T t )
T
xt + y (T t )
Tx
(a)
(b)
(29)
t
T
for duration t
(a)
(b)
x 2 Rt + y 2 R ( T t )
T
(c)
Average loss =
(30)
Rx 2 t + y 2 R ( T t ) t y T t
= +
= FLS =
RTx 2
T x T
2
Loss factor
(d)
T t
1. 0
T
t
2
y
0 and x 0 FLS (FLD)
T
= Fp
= cos ( - ) = cos
(33)
= Active Power, P
r
Apparent Power, S
In equation (33), = arc tan
Q
. Here again this is true only for pure sinusoidal qualities.
P
industrial and commercial services, the contracts of which might contain power factor
clauses.
Balanced load: A balanced polyphase load is that which draws symmetrical currents when
symmetrical voltages are applied to it.
Balanced voltages = symmetrical polyphase voltages.
Voltage unbalance = maximum deviation from the average base phase voltage/average
phase voltage.
(34)
or V2 =
V1
negative-sequence voltage
positive-sequence voltage
(35)
This could easily be achieved more easily with the aid of computers.
The second definition is widely used. We will get more into it when we discuss the
subject of symmetrical components and overcurrent protection.
Load Distribution and Load Density
2 methods of representing distributed load
a) uniform distribution
b) loads are all same fraction of distribution transformer rating
2 methods of load density representation
a) linear density - kVA/1000
b) area density - MVA/mi2
Determining Load Characteristics
1) Measurements
2) Statistics (using a few spot measurements)
The greatest interest is in residential loads as opposed to commercial or industrial.
The second method uses factors to estimate maximum demand for each consumer.
The factors are:
0.0039 for 0-500 kWh
0.0035 for 500-1500 kWh
0.0030 for > 1500 kWh
Coincidence or Diversity Factor
1 Fc
FcN
Fc
FcN
Fc
FcN
DN
D1
DN
(51)
loads
loads
(52)
D1
Fc
nFcn1
n 1
(53)
DN
D1FcN
(54)
EXAMPLE:
Usually such transformers serve 4 to 14 houses depending upon the type of heating and
transformer sizes. Further, let us consider that each house has a clothes dryer, a range, a
refrigerator, lighting and miscellaneous loads. We want to estimate the 30-minute
maximum diversified demand as seen by the transformer. We are also given that the
maximum coincident demand seen by each of these load types for each house is (see
Figure 7)
Dryer
1.6 kW
Range:
0.8 kW
distribution transformer
Refrigerator: 0.0666 kW
Lighting:
6 similar houses
0.61 kW
18.5 kW
curve,
1.2 kW/house
0.53 kW/ house
range
0.52 kW/house
refrigerator
0.044 kW/house
Pav, max =
dryer
lighting, etc.
clothes dryer
12am
1 am
2 am
3 am
4 am
5 am
6 am
7 am
8 am
9 am
0.03
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.08
0.20
1.6 x 6 kW
Range 0.24
0.8 x 6 kW
Refrigerator 0.90 x
0.066 x 6 kW
lighting 0.32
0.61 x 6 kW
10 am
11 am
12 noon
1 pm
2 pm
3 pm
4 pm
0.65
1.00
0.98
0.70
0.65
0.63
0.38
Table 1
Hourly variation factors
Flat Demand Rate: Does not require meters. Used for street lighting.
2.
3.
4.
Demand Rate: Recognizes load factor and involves a two-part tariff, one based on
maximum (monthly) demand and the second one based on actual energy
consumed.
5.
Seasonal Rate: Summer versus Winter Rate. Here in the Northwest we pay
higher rate for winter and lower rate for summer. (Why?)
6.
Load Growth
P0(1 + g)n
Pn
Pn:
(load forecasting)
g:
n:
# of years
P0 :
initial load
Pn
P0
n
0
5. Rate Structure
Sample:
Min change (first 20 kWh)
Next 80 kWh
$0.0355/kWh
$0.0321/kWh
$0.0296/kWh
$0.0265/kWh
$2.25/mo.
$0.0220/kWh
Consumption 2200kWh
2.25 + 80 x 0.0355 + 100 x 0.0321 + 200 x 0.0296 + 400 x 0.0265 + 1400 x 0.022
=
55.62
Environmental surcharges:
County tax, fuel cost adjustment, state sales tax
Total: $ 85.42
*rate schedule may be seasonal.
Northwest higher rate in Winter higher cost due to power import from California.
Fuel Cost Adjustment (factor FCAF)
FCAF =
$/kWh
Btu
kWh
B
10 6
1
1 D
$
Btu
A:
B:
C:
KWH x
KWH x
1
:
total generation to meet the given demand
1 D
loss factor
1
C :
1 D