Load Flow Modelling Service Project results 30 September 2002

1
Load Flow Modelling ServiceProject results 30
September 2002

• Mick Barlow
• Srdjan Curcic

2
Content

• Project objectives
• Modelling assumptions
• Some potential issues with the proposed
  approaches to allocation of losses
• Illustration of the key results

3
Project objectives

• Power Technologies Int. has been commissioned to
  assist the assessment procedure of modification
  proposals P75 and P82, specifically
• to perform calculations of TLFs TLMs for a
  specified number of SPs for P75 P82
• to present the results in a form suitable for the
  assessment procedure and
• to draw attention to potential issues with the
  fundamentals of the two marginal approaches
  proposed (P75 P82)
• This presentation contains an appropriate
  selection of the project results

4
Modelling assumptions

• MW injections are calculated from the metered
  volumes assuming they are constant (average)
• Power factors published in NGCs SYS are used to
  calculate reactive power injections
• It is assumed that offtake metered
  volumes/injections are accurate and then the
  delivery volumes/injections are calculated to
  balance offtake metered volumes/injections and
  calculated variable active power system losses,
  while maintaining relative deliveries among the
  generators (DICTATED BY DATA QUALITY)

5
Modelling assumptions

• Load flow assumptions
• Generation P is calculated as described above
• Generation Q is calculated automatically by
  setting voltage target to 1.03 p.u.
• Load P is calculated from the metered volumes
• Load Q is calculated from typical power factors
  given in the NGCs SYS
• Transformer tap changer are set according to
  information in the NGCs SYS for each voltage
  level
• Transformer target voltage is set to produce
  reasonable voltage profile (between 0.97 and 1.03
  p.u.)
• SVCs target voltage is set to values recommended
  in the NGCs SYS

6
Modelling assumptions

• AC load flow calculations are using the standard
  NGCs slack at Cowley
• PTIs PSS/E-OPF is used for calculating marginal
  TLFs
• Out of TLFs obtained for active power injections
  and reactive power injections, unique TLFs are
  calculated that relate only to active power
  injections, while providing for the total losses
  incurred from a node
• TLFij (TLFPij ? Pij TLFQij ? Qij) / Pij

7
Modelling assumptions

• On the basis of information in the NGCs SYS, the
  fixed losses are assumed to be 200MW (peak),
  180MW (trough) and 190MW (other periods Autumn)
  
• For the purpose of calculating TLMs, these fixed
  losses are smeared across generators,
  proportionally to their power output

8
Issues with proposed allocation of losses

• Slack node An issue with the marginal TLFs
  approach

• The choice of the slack node potentially matters
  more that initially expected

9
Issues with proposed allocation of losses
Slack node
Network metered volumes are for 22 January
2002 Comparison is between slack node at Cowley
and at Thorpe Marsh
Illustrative example Assume a generation metered
volume of 1000MWh and a TLM of 0.94. That would
attribute 60MWh of losses to this generation.
Due to the indicated change in slack node this
60MWh of losses would change for 9.23, on
average.
10
Issues with proposed allocation of losses
Power factor

• The sensitivity analysis to introduction/variation
  of power factors has been done in an simple
  exercise
• For 02 January 2002 network and metered volumes
  two cases were calculated
• P.F. 1 for all demand nodes
• P.F. by NGCs SYS

11
Issues with proposed allocation of losses
Power factor

• At the level of TLFs the effect was much more
  tangible. This indicates
• that reactive powers should not be ignored and
• a possible need for further consideration

12
Summary Issues with proposed allocation of losses

• There are some issues arising from the modelling
• TLFMG, nevertheless, have confidence in the
  modelling results

13
Illustration of key resultsP75
14
P75 (T1)Base cases02 Jan. 02 (SP36) - Peak01
Aug. 01 (SP8) - Trough10 Oct. 01 (SP25) Week
day daylight11 Oct. 01 (SP11) Week day night
13 Oct. 01 (SP25) Weekend daylight 14 Oct. 01
(SP11) Weekend night
15
GSPG and TNUoS (gen) zones
16
Change P75 introduces an example (based on
marginal, GSPG zone, ½ h TLFs)
02 January 2002
North
South
17
Demand TLMs for P75 (based on marginal, GSPG
zone, ½ h TLFs)
North
South
18
Change P75 introduces an example (based on
marginal, TNUoS (gen) zone, ½ h TLFs)
02 January 2002
North
South
19
Generation TLMs for P75 (based on marginal, TNUoS
(gen) zone, ½ h TLFs)
North
South
20
P75 (T1) - Summary

• Introduction of Modification Proposal P75 would
  result in variable TLMs
• over time and
• across country
• with an overall, indicative variation between
  0.95 and 1.06 for demand and between 0.94 and
  1.09 for generation

21
P75 (T2)Variation of TLMs over time 02 Jan. 02
(SP8 SP36) Peak day01 Aug. 01 (SP8 SP36)
Trough day 10 Oct. 01 (SP1 SP48) Week day11
Oct. 01 (SP11) Week day night 13 Oct. 01
(SP25) Weekend daylight 14 Oct. 01 (SP11)
Weekend night
22
P75 GSPG zone TLMs over sample time period -
demand
North
South
23
P75 TNUoS (gen) zone TLMs over sample time
period - generation
North
South
24
P75 GSPG zone TLMs over a day - Demand
10 October 2001
South
North
25
P75 TNUoS (gen) zone TLMs over a day - Generation
10 October 2001
South
North
26
P75 (T2) - Summary

• Introduction of Modification Proposal P75 would
  result in daily variation in TLMs of up to
  approximately 0.03 for demand and 0.045 for
  generation (the exception is TNUoS zone 5 with
  reversible hydro plants where the variation is up
  to 0.065) on a typical autumn working day

27
P75 (T3)Sensitivity to network
configurationIndicative/Intact/Representativenet
works for the following SPs 02 Jan. 02 (SP36) -
Peak01 Aug. 01 (SP8) - Trough10 Oct. 01 (SP25)
Week day daylight11 Oct. 01 (SP11) Week day
night 13 Oct. 01 (SP25) Weekend daylight 14
Oct. 01 (SP11) Weekend night
28
P75 GSPG zone TLMs - Demandsensitivity to
network configuration
02 January 2002 (peak)
North
South
29
P75 TNUoS (gen) zone TLMs - Generationsensitivit
y to network configuration
02 January 2002 (peak)
North
South
30
P75 GSPG zone TLMs - Demandsensitivity to
network configuration
01 August 2002 (trough)
North
South
31
P75 TNUoS (gen) zone TLMs - Generationsensitivit
y to network configuration
01 August 2002 (trough)
North
South
32
P75 GSPG zone TLMs - Demandsensitivity to
network configuration
10 October 2001 (weekday daylight)
North
South
33
P75 TNUoS (gen) zone TLMs - Generationsensitivit
y to network configuration
10 October 2001 (weekday daylight)
North
South
34
P75 GSPG zone TLMs - Demandsensitivity to
network configuration
11 October 2001 (weekday night)
North
South
35
P75 TNUoS (gen) zone TLMs - Generationsensitivit
y to network configuration
11 October 2001 (weekday night)
North
South
36
P75 GSPG zone TLMs - Demandsensitivity to
network configuration
13 October 2001 (weekend daylight)
North
South
37
P75 TNUoS (gen) zone TLMs - Generationsensitivit
y to network configuration
13 October 2001 (weekend daylight)
North
South
38
P75 GSPG zone TLMs - Demandsensitivity to
network configuration
14 October 2001 (weekend night)
North
South
39
P75 TNUoS (gen) zone TLMs - Generationsensitivit
y to network configuration
14 October 2001 (weekend night)
North
South
40
P75 (T3) - Summary

• Network configuration can have an effect on TLMs
• while there is almost no difference between
  intact and representative networks,
• there is a tangible difference between indicative
  and intact networks

41
P75 (T4)Sensitivities to constraints02 January
2002 (base case)02 January 2002 (constrained
case)
5 double circuits restricted for 20 below the
flow level in base case
42
P75 GSPG zone TLMs - Demandsensitivity to
constraints
Losses 758.5MW (base case), 573 (constrained
case)
North
South
43
P75 TNUoS zone TLMs - Generation sensitivity to
constraints
Losses 758.5MW (base case), 573 (constrained
case)
North
South
44
P75 (T4) - Summary

• Constraints may have an impact on TLMs

45
P75 (T5)Comparison of Generation TLFs/TLMs and
Demand TLFs/TLMs at the same node
46
P75 Comparison of Generation/Demand Zonal
TLFs/TLMs
Node Rye House GSPG zone 7 TNUoS
zone 10 SP 02 January 2002, SP36
47
P75 (T5) - Summary

• Discrepancies between TLMs for generation and
  demand at a node are not greatly exacerbated if
  generation and demand zones are different from
  one another
• Only one node has been assessed and this may not
  be representative

48
P75 (T6)Comparison of Nodal TLFs/TLMs with Zonal
TLFs/TLMs
49
P75 Comparison of Nodal/Zonal TLMs (Demand)
02 January 2002
Winter peak (SP36)
50
P75 Comparison of Nodal/Zonal TLMs (Generation)
02 January 2002
Winter peak (SP36)
51
P75 Comparison of Nodal/Zonal TLMs (Demand)
01 August 2001
Summer trough (SP8)
52
P75 Comparison of Nodal/Zonal TLMs (Generation)
01 August 2001
Summer trough (SP8)
53
P75 (T6) - Summary

• Introduction of Modification Proposal P75 and its
  zonal TLMs (GSPG zones for demand and TNUoS (gen)
  zones for generation) would result in nodal TLMs
  for some nodes being closer to neighbouring zonal
  TLMs

54
P75 (T7)Model performance characteristics
55
P75 Input data

• Data capture depends on the source systems
• Data capture will be greatly improved once NGCs
  state estimator is fully operational
• Amount, duration and quality of data depends on
  the source systems and the TLF modification
  applied
• The quality and consistency of the data determine
  the time and effort required to establish
  acceptable loadflow conditions

56
P75 Processing and output data

• Once a suitable loadflow data set has been
  obtained it should take less than 10 seconds to
  obtain TLFs for all considered generation/demand
  points for a SP
• The output would be suitably structured list of
  TLFs
• Further post processing could be established to
  automatically produce a higher level of the
  output data

57
P75 (T8)Sensitivities to flows on French
Inter-connector31 October 2001 (importing)07
November 2001 (exporting)
58
P75 GSPG zone TLMs -sensitivity to flows on
French Inter-connector
Losses 377.1MW (importing), 533.3 (exporting)
North
South
59
P75 TNUoS zone TLMs -sensitivity to flows on
French Inter-connector
Losses 377.1MW (importing), 533.3 (exporting)
North
South
60
P75 (T910)Examples of moving generation
deliveries for the selected nodes
Greystones
Barking Power
61
P75 GSPG zone TLMs -sensitivity to plant
outages / response to signals
02 January 2002
North
South
62
P75 TNUoS (gen) zone TLMs -sensitivity to plant
outages / response to signals
02 January 2002
North
South
63
For the base case of 02 January 2002 500MW were
moved from Dinorwig to 5 different places
P75 (T11) Sensitivity to intermittent generation
Blyth
Harker
Norwich
Dinorwig
Grain
Indian Queens
64
P75 GSPG zone TLMs - Demandsensitivity to
intermittent generation
02 January 2002
North
South
65
P75 TNUoS zone TLMs - Generationsensitivity to
intermittent generation
02 January 2002
North
South
66
P75 (T8 to T11) - Summary

• TLMs in Modification Proposal P75 are only
  locally sensitive to plant withdrawal or plant
  relocation or to intermittent generation
• Note that for Task 8 (sensitivity to flow on
  French inter-connector) the effect is complicated
  as we have two different dates and SPs with
  different generation and demand profiles

67
Illustration of key resultsP82
68
P82 (T1)Variation of scaled TLFs over time02
Jan. 02 (SP8 SP36) Peak day01 Aug. 01 (SP8
SP36) Trough day 10 Oct. 01 (SP1 SP48)
Week day11 Oct. 01 (SP11) Week day night 13
Oct. 01 (SP25) Weekend daylight 14 Oct. 01
(SP11) Weekend night
69
P82 GSPG zone ½ h TLMs over sample time period -
Demand
North
South
70
P82 GSPG zone ½ h TLMs over sample time period -
Generation
North
South
71
P82 GSPG zone ½ h TLMs over a day - Demand
10 October 2001
South
North
72
P82 GSPG zone ½ h TLMs over a day - Generation
10 October 2001
South
North
73
P82 (T1) - Summary

• Before implementing the averaging of TLFs, the
  variation in TLMs for P82 would be up to
  approximately 0.016 for demand and 0.017 for
  generation on a typical autumn working day

74
P82 (T2)Base cases 02 Jan. 02 (SP36) - Peak01
Aug. 01 (SP8) - Trough10 Oct. 01 (SP25) Week
day daylight11 Oct. 01 (SP11) Week day night
13 Oct. 01 (SP25) Weekend daylight 14 Oct. 01
(SP11) Weekend night
75
GSPG and TNUoS (gen) zones
76
Change P82 introduces an example (based on
scaled, GSPG zone, averaged TLFs)
02 January 2002
North
South
77
Demand TLMs for P82 (based on scaled, GSPG zone,
averaged TLFs)
North
South
78
Change P82 introduces an example (based on
scaled, GSPG zone, averaged TLFs)
02 January 2002
North
South
79
Generation TLMs for P82 (based on scaled, GSPG
zone, averaged TLFs)
North
South
80
P82 (T2) - Summary

• Introduction of Modification Proposal P82 would
  result in variable TLMs
• over time and
• across country
• with an overall variation between 0.99 and 1.03
  for demand and between 0.975 and 1.015 for
  generation, where spatial variations are larger
  than temporal variations

81
P82 (T3)Variation of TLMs over time(based on
average scaled TLFs)02 Jan. 02 (SP8 SP36)
Peak day01 Aug. 01 (SP8 SP36) Trough day 10
Oct. 01 (SP1 SP48) Week day11 Oct. 01 (SP11)
Week day night 13 Oct. 01 (SP25) Weekend
daylight 14 Oct. 01 (SP11) Weekend night
82
P82 GSPG zone TLMs over sample time period -
Demand
North
South
83
P82 GSPG zone TLMs over sample time period -
Generation
North
South
84
P82 GSPG zone TLMs over a day - Demand
10 October 2001
South
North
85
P82 GSPG zone TLMs over a day - Generation
10 October 2001
South
North
86
P82 (T3) - Summary

• Introduction of Modification Proposal P82 would
  result in almost non-existent daily variations in
  TLMs (up to approximately 0.004 for demand and
  0.004 for generation) on a typical autumn working
  day

87
P82 (T4)Sensitivity of TLFs TLMs to choice of
historic data and weightings
88
P82 Sensitivity of TLFs TLMs to historic data
and weighting coefficients

• This task has not been completed as the required
  input data have not been provided

89
P82 (T5)Sensitivities to constraints02 January
2002 (base case)02 January 2002 (constrained
case)
5 double circuits restricted for 20 below the
flow level in base case
90
P82 GSPG zone TLMs - Demandsensitivity to
constraints
Losses 758.5MW (base case), 573 (constrained
case)
North
South
91
P82 TNUoS zone TLMs - Generation sensitivity to
constraints
Losses 758.5MW (base case), 573 (constrained
case)
North
South
92
P82 (T5) - Summary

• TLMs in Modification Proposal P82 are almost
  insensitive to constraints because the effects
  of such events would be averaged in the following
  year

93
P82 (T6)Comparison of Nodal TLFs/TLMs with Zonal
TLFs/TLMs
94
P82 Comparison of Nodal/Zonal TLMs Generation
02 January 2002
Winter peak (SP36)
95
P82 Comparison of Nodal/Zonal TLMs - Demand
02 January 2002
Winter peak (SP36)
96
P82 Comparison of Nodal/Zonal TLMs Generation
01 August 2001
Summer trough (SP8)
97
P82 Comparison of Nodal/Zonal TLMs - Demand
01 August 2001
Summer trough (SP8)
98
P82 (T6) - Summary

• Introduction of Modification Proposal P82 and its
  zonal TLMs (GSPG zones only for both demand and
  generation) would result in nodal TLMs for some
  nodes being closer to neighbouring zonal TLMs

99
P82 (T7)Sensitivities to flows on French
Inter-connector31 October 2001 (importing)07
November 2001 (exporting)
100
P82 GSPG zone TLMs - Demandsensitivity to flows
on French Inter-connector
Losses 377.1MW (importing), 533.3 (exporting)
North
South
101
P82 GSPG zone TLMs - Generationsensitivity to
flows on French Inter-connector
Losses 377.1MW (importing), 533.3 (exporting)
North
South
102
P82 (T7) - Summary

• TLMs in Modification Proposal P82 are almost
  insensitive to direction of flows on the French
  inter-connector because the effects of such
  events would be averaged in the following year

103
P82 (T8) Degree to which a scaling factor of 0.5
recovers heating losses 02 Jan. 02 (SP36) -
Peak01 Aug. 01 (SP8) - Trough10 Oct. 01 (SP25)
Week day daylight11 Oct. 01 (SP11) Week day
night 13 Oct. 01 (SP25) Weekend daylight 14
Oct. 01 (SP11) Weekend night
104
P82 Comparison between flat 0.5 scaling factor
and the actual reconciliation factors
105
P82 Nodal TLFs recovery of heating losses with
scaling factor 0.5 and the actual reconciliation
factors
106
P82 (T8) - Summary

• Scaling factor of 0.5 does not precisely recover
  heating losses but appears to be a reasonable
  approximation

107
P82 (T910)Examples of moving generation
deliveries for the selected nodes
Greystones
Barking Power
108
P82 GSPG zone TLMs - Demand sensitivity to
plant outages / response to signals
02 January 2002
North
South
109
P82 GSPG zone TLMs - Generation sensitivity to
plant outages / response to signals
02 January 2002
North
South
110
For the base case of 02 January 2002 500MW were
moved from Dinorwig to 5 different places
P82 (T11) Sensitivity to intermittent generation
Blyth
Harker
Norwich
Dinorwig
Grain
Indian Queens
111
P82 GSPG zone TLMs - Demandsensitivity to
intermittent generation
02 January 2002
North
South
112
P82 GSPG zone TLMs - Generationsensitivity to
intermittent generation
02 January 2002
North
South
113
P82 (T9 to T11) - Summary

• TLMs in Modification Proposal P82 are almost
  insensitive to plant withdrawal or move or to
  intermittent generation because the effects of
  such events would be averaged in the following
  year

114
END OF PRESENTATION