Advanced Search

Load Flow Model Specification for the Calculation of Nodal Transmission Loss Factors

v 3.0
Download

Load Flow Model Specification for the Calculation of Nodal Transmission Loss Factors

Version 3.0

Effective Date: 29 March 2019

AMENDMENT HISTORY

Implementation Date

Version

Description of Change

Mods/ Panel/ Committee Refs

June 2003

1.0

Baseline version for P82

CVA Programme

12 May 2017

2.0

Modification P350

Panel 266/10

29 March 2019

3.0

29 March 2019 Standalone Release for P369

Panel 285/12

Intellectual Property Rights, Copyright and Disclaimer

The copyright and other intellectual property rights in this document are vested in ELEXON or appear with the consent of the copyright owner. These materials are made available for you for the purposes of your participation in the electricity industry. If you have an interest in the electricity industry, you may view, download, copy, distribute, modify, transmit, publish, sell or create derivative works (in whatever format) from this document or in other cases use for personal academic or other non-commercial purposes. All copyright and other proprietary notices contained in the document must be retained on any copy you make.

All other rights of the copyright owner not expressly dealt with above are reserved.

No representation, warranty or guarantee is made that the information in this document is accurate or complete. While care is taken in the collection and provision of this information, ELEXON Limited shall not be liable for any errors, omissions, misstatements or mistakes in any information or damages resulting from the use of this information or action taken in reliance on it.

1 Introduction

This Load Flow Model (LFM) Specification has been established by the Panel in accordance with Balancing and Settlement Code (BSC) Section T Annex T-2 paragraphs 2 and 3, with support from BSCCo and the National Electricity Transmission System Operator (NETSO). This LFM Specification is a Code Subsidiary Document, which forms part of the Service Description of the Transmission Loss Factor Agent (TLFA) (reference 1).

The LFM Specification contains the requirements, obligations, assumptions and approximations required to be supported by the LFM. The exact mechanism for the derivation of Nodal Transmission Loss Factors (TLFs) by the TLFA is the required function of the LFM. For the avoidance of doubt, the LFM produces Nodal TLFs and any further data manipulation is carried out by the Transmission Loss Factor Agent. For example, converting the Nodal TLFs into Zonal TLFs and then into BM Unit specific TLFs.

In the event of any discrepancy between the LFM Specification and the TLFA Service Description, Section H 1.5.2 (b) of the Code places the obligation on the Panel, with support from BSCCo, to determine the precedence and resolve any discrepancy by raising the relevant amendment in accordance with Section F.3. of the Code. Furthermore, in the event of any discrepancy between the LFM Specification and the Code Section H 1.5.2 (b) places the obligation on the Panel, with support from BSCCo, to determine precedence, and to resolve the discrepancy by raising the relevant amendment, in accordance with the Code, Section F 3.

1.1 Model Reviewer

The BSC Panel will appoint a Model Reviewer who will verify that the LFM produced by the TLFA produces Nodal TLFs in accordance with this LFM Specification, in accordance with Section T, Annex T-2 of the Code. Load Flow Model

1.2 Background

The intent of Load Flow Model is to derive a set of annual average Transmission Loss Factors (TLFs) to recover heating losses on a zonal basis and fixed losses on a uniform basis, using a scaling factor of 0.5. The TLFs are to be derived annually on an ex ante basis using historical metered and network data. Nodal TLFs will be derived for a set of Sample Settlement Periods.

A Load Flow Model to be used for evaluation of TLFs is to be based on a DC load flow, i.e. a modelling approach for an interconnected network utilising data reflective of alternating current (AC) electrical flows on that network, but with a set of simplifying assumptions that render the equations for the AC flows similar in form to those for a direct current (DC) flow. The Load Flow Model Produces Nodal TLFs for each Sample Settlement Period.

1.3 Objectives

A LFM is a mathematical model of an electrical network which represents power flows between pairs of adjacent nodes on the network, and from which Nodal TLFs can be determined for each Node for given power flows. TLFs are representative of the changes in transmission losses arising from marginal changes in demand or generation at Nodes on the Transmission Network.

The key objectives of the LFM are to:

1. Accurately represent the physical characteristics of the England and Wales Transmission Network via a direct current (DC) load flow model;

2. Use Network Data that reflects, as far as is reasonably possible, the conditions prevailing on the network at any time, representative of an ‘intact network’, i.e. a complete England and Wales Transmission Network assuming no circuits de-energised or disconnected with all lines in operation;

3. Capture the delivery, injections onto the network, and offtake, withdrawals from the network, for a large number of Nodes for Sample Settlement Periods throughout each Reference Year; and

4. Generate TLFs that are representative of the changes in transmission losses arising from marginal changes in demand or generation at nodes on the Transmission Network.

1.4 Assumptions and Approximations

The Load Flow Model Specification shall provide for the following assumptions and approximations to be made in the Load Flow Model:

1. Only electrical losses associated with power flows on circuits (forming part of the network) will be used in determining Nodal TLFs (fixed losses will be set in line with those in the NETSO Seven Year Statement)

2. In respect of the power flow between adjacent nodes, it is assumed that:

    1. There is no Reactive Power component;

    2. The ratio of the change of power flow over a circuit to the injection at a given node is not dependent on overall electrical load on the network;

    3. The sine of the voltage phase angle is equal to the phase angle (as measured in radians); and

    4. The power flow in a circuit is equal to the difference in the voltage phase angles across the circuit multiplied by the circuit susceptance.

2 Load Flow Model Requirements

On the basis of the required assumptions listed above, the specification of an appropriate DC load flow model is presented in the following sections (3.1 to 3.3). A conventional DC Load Flow Model relates real power flows (i.e. generation or demand MW) to voltage phase angle (voltage magnitude being assumed constant and equal to 1 pu) using only branch reactances, all resistance being ignored.

The process of computing TLFs based on such a DC Load Flow Model will involve the following three steps:

    • STEP 1: Calculate adjusted nodal power flows from Nodal metered generation and demand data, suitable for the application of the conventional DC Load Flow Model;

    • STEP 2: Calculate network power flows using the conventional DC Load Flow Model;

    • STEP 3: Determine flow-injections sensitivity factors and compute TLFs

These steps are detailed in the following section 3.1 to 3.3 and an example is contained in Appendix 2.

2.1 STEP 1: Calculation of adjusted nodal power flows from metered generation and demand data

The conventional DC Load Flow Model excludes consideration of losses in the process of evaluating voltage phase angles and flows. It is proposed that a simple adjustment of metered volumes of generation (MWh) and demand (MWh) is performed and used to compute Nodal power flows as the input to the DC Load Flow Model:

Gn=G¯ n(1L2G¯ ) size 12{G rSub { size 8{n} } = {overline {G}} rSub { size 8{ {} rSub { size 6{n} } } } \( 1 - { {L} over { size 12{2 Sum { {overline {G}} } } } } size 12{ \) }} {} and (1)

Dn=D¯ n(1+L2D¯ ) size 12{D rSub { size 8{n} } = {overline {D}} rSub { size 8{ {} rSub { size 6{n} } } } \( 1+ { {L} over { size 12{2 Sum { {overline {D}} } } } } size 12{ \) }} {} (2)

where

G¯ n size 12{ {overline {G}} rSub { size 8{n} } } {} and D¯ n size 12{ {overline {D}} rSub { size 8{n} } } {} are metered generation and demand respectively at Node n, and where L is the metered losses calculated as follows:

L=Gn¯ D¯ n size 12{L= Sum { {overline {G rSub { size 8{n} } }} } - Sum { {overline {D}} rSub { size 8{n} } } } {} (3)

Gn and Dn are Nodal power flows to be used in the DC Load Flow Model

This adjustment allows the conventional (loss-inclusive) DC Load Flow Model to be applied for the evaluation of network power flows since:

GnDn=0 size 12{ Sum cSub {} {G rSub { size 8{n} } - {}} Sum cSub {} {D rSub { size 8{n} } } =0} {} (4)

Note that this process will produce consistent inputs for the DC Load Flow even if the metered data is inconsistent. For example, in case that the metered losses are inconsistent with metered generation and demand, as well as in the extreme case of the total metered generation being smaller than the total metered demand. The example presented in the Appendix 2 illustrates the adequacy of the proposed approach to computing Nodal power flows from the metered data.

2.2 STEP 2: Evaluation of network power flows using the conventional DC Load Flow Model

Active power balance at each of the Nodes is given by the following expression:

GaDa=Pa=n=1N|Ea||En|[Gancos(θaθn)+Bansin(θaθn)] size 12{G rSub { size 8{a} } - D rSub { size 8{a} } =P rSub { size 8{a} } = Sum cSub { size 8{n=1} } cSup { size 8{N} } { lline E rSub { size 8{a} } rline lline E rSub { size 8{n} } rline left [G rSub { size 8{ ital "an"} } "cos" \( θ rSub { size 8{a} } - θ rSub { size 8{n} } \) +B rSub { size 8{ ital "an"} } "sin" \( θ rSub { size 8{a} } - θ rSub { size 8{n} } \) right ]} } {} a = 1,..,n (5)

Where

Pa=GaDa size 12{P rSub { size 8{a} } =G rSub { size 8{a} } - D rSub { size 8{a} } } {}The net Nodal power flow, defined as the difference between generation and demand at the corresponding Node

θa,θn size 12{θ rSub { size 8{a} } ,`θ rSub { size 8{n} } } {} the phase angles at Nodes a and n respectively

|Ea|,|En| size 12{ lline E rSub { size 8{a} } rline , lline E rSub { size 8{n} } rline } {} the voltage magnitudes at Nodes a and n respectively

Gan+jBan=Yan size 12{G rSub { size 8{ ital "an"} } + ital "jB" rSub { size 8{ ital "an"} } =Y rSub { size 8{ ital "an"} } } {} - the a-n term in the complex Y matrix of the power network

The conventional DC load flow is obtained by

      1. neglecting losses in power flow calculations Gab=0 size 12{G rSub { size 8{ ital "ab"} } =0} {},

      2. assuming that the voltage magnitudes at all Nodes equal to 1 p.u (|En|=1 size 12{ lline E rSub { size 8{n} } rline =1} {}).

      3. assuming that the sine of the voltage phase angle is equal to the phase angle:

sin(θaθn)(θaθn) size 12{"sin" left (θ rSub { size 8{a} } - θ rSub { size 8{n} } right ) approx left (θ rSub { size 8{a} } - θ rSub { size 8{n} } right )} {}

The corresponding load flow equations constitute a DC power flow:

Pa=n=1NBan(θaθn) size 12{P rSub { size 8{a} } = Sum cSub { size 8{n=1} } cSup { size 8{N} } {B rSub { size 8{ ital "an"} } \( θ rSub { size 8{a} } - θ rSub { size 8{n} } \) } } {} (6)

given that

Bab=1xan size 12{B rSub { size 8{ ital "ab"} } = { {1} over {x rSub { size 8{ ital "an"} } } } } {} (7)

where, xan size 12{x rSub { size 8{ ital "an"} } } {}is the reactance between Nodes a and n, the corresponding conventional DC Load Flow Model can be presented the standard matrix form:

[P1P2...PN]=[n11x1n1x12...1x1N1x21n21x2n...1x2N............1xn1......nN1xNn][θ1θ2...θN] size 12{ left [ matrix { P rSub { size 8{1} } {} ## P rSub { size 8{2} } {} ## "." "." "." {} ## P rSub { size 8{N} } } right ]= left [ matrix { Sum cSub { size 8{n <> 1} } { { {1} over {x rSub { size 8{1n} } } } } {} # - { {1} over {x rSub { size 8{"12"} } } } {} # "." "." "." {} # - { {1} over {x rSub { size 8{1N} } } } {} ## - { {1} over {x rSub { size 8{"21"} } } } {} # Sum cSub { size 8{n <> 2} } { { {1} over {x rSub { size 8{2n} } } } } {} # "." "." "." {} # - { {1} over {x rSub { size 8{2N} } } } {} ## "." "." "." {} # "." "." "." {} # "." "." "." {} # "." "." "." {} ## - { {1} over {x rSub { size 8{n1} } } } {} # "." "." "." {} # "." "." "." {} # Sum cSub { size 8{n <> N} } { { {1} over {x rSub { size 8{ ital "Nn"} } } } } {} } right ] cdot left [ matrix { θ rSub { size 8{1} } {} ## θ rSub { size 8{2} } {} ## "." "." "." {} ## θ rSub { size 8{N} } } right ]} {} (8)

where:

P1,..,PN represents net power flow injections (given) at Nodes 1 to N,

θ1 size 12{θ rSub { size 8{1} } } {},…,θN size 12{θ rSub { size 8{N} } } {} voltage phase angles (to be calculated) at Nodes 1 to N, and

xab size 12{x rSub { size 8{ ital "ab"} } } {} is the reactance of the circuits between Nodes a and b (given).

The net power flow is defined as the difference between generation and demand at the corresponding Node (Pn=GnDn size 12{P rSub { size 8{n} } =G rSub { size 8{n} } - D rSub { size 8{n} } } {}).

The matrix representing network characteristics (both the topology and electrical parameters of the circuits - reactances), belongs to the class of so-called Ybus matrices, and is presented in (9). The diagonal elements of the matrix correspond to the sum of susceptances coincident with the corresponding Node, while off diagonal elements correspond to the negative values of susceptance linking the corresponding Nodes.

[Ybus]=[n11x1n1x12...1x1N1x21n21x2n...1x2N............1xn1......nN1xNn] size 12{ left [Y rSub { size 8{ ital "bus"} } right ]= left [ matrix { Sum cSub { size 8{n <> 1} } { { {1} over {x rSub { size 8{1n} } } } } {} # - { {1} over {x rSub { size 8{"12"} } } } {} # "." "." "." {} # - { {1} over {x rSub { size 8{1N} } } } {} ## - { {1} over {x rSub { size 8{"21"} } } } {} # Sum cSub { size 8{n <> 2} } { { {1} over {x rSub { size 8{2n} } } } } {} # "." "." "." {} # - { {1} over {x rSub { size 8{2N} } } } {} ## "." "." "." {} # "." "." "." {} # "." "." "." {} # "." "." "." {} ## - { {1} over {x rSub { size 8{n1} } } } {} # "." "." "." {} # "." "." "." {} # Sum cSub { size 8{n <> N} } { { {1} over {x rSub { size 8{ ital "Nn"} } } } } {} } right ]} {} (9)

In order to solve system of equations (8) a reference slack node needs to be chosen, since (9) is a singular matrix and hence equations (8) are linearly dependent. With no loss of generality but for the sake of simplicity of the presentation, Node 1 is declared as the slack node. The system of equation (8) can be now solved and the corresponding voltage phase angles determined using matrix techniques routinely applied in load flow calculations:

[θ]=[000[Yr]1][P] size 12{ left [θ right ]= left [ matrix { 0 {} # 0 {} ## 0 {} # left [Y rSub { size 8{r} } right ] rSup { size 8{ - 1} } {} } right ]` left [P right ]} {} (10)

where Yr is obtained by removing the row and the column from the Ybus that correspond to the slack node.

Once the voltage phase angles are calculated (10), circuit flows can be computed:

Fk=1xab(θaθb) size 12{F rSub { size 8{k} } = { {1} over {x rSub { size 8{ ital "ab"} } } } \( θ rSub { size 8{a} } - θ rSub { size 8{b} } \) } {} (11)

where Fk is the power flow in a circuit k, and circuit k is between Nodes a and b.

2.3 STEP 3: Determine power injection sensitivity factors and compute TLFs

A Nodal TLF, associated with a particular Node n, is defined as the incremental change in the network losses (L) due to an incremental increase in power injection (Pn) at Node n:

TLFn=ΔLΔPn size 12{ ital "TLF" rSub { size 8{n} } = { {ΔL} over {ΔP rSub { size 8{n} } } } } {} (12)

(Symbol Δ size 12{Δ} {} indicates an incremental change)

As indicated above, network losses will be divided in “heating losses”, which depend on network loading conditions, and “fixed losses” that are independent from network loading. Therefore, the network model to be used for Nodal TLF evaluations will only include components that generate heating losses, which means that the network model will contain only series impedances and exclude all shunt impedances.

In a network with the total number of circuits (network branches) being M, the total “heating losses” are the sum of losses attributed to each individual transmission circuits k in the network:

Lh=k=1MLkh size 12{L rSup { size 8{h} } = Sum cSub { size 8{k=1} } cSup { size 8{M} } {L rSub { size 8{k} } rSup { size 8{h} } } } {} (13)

The Nodal TLF associated with Node n can now be expressed as follows:

TLFn=ΔLhΔPn=ΔΔPn(k=1MLkh)=k=1McΔLkhΔPn size 12{ ital "TLF" rSub { size 8{n} } = { {ΔL rSup { size 8{h} } } over {ΔP rSub { size 8{n} } } } = { {Δ} over {ΔP rSub { size 8{n} } } } \( Sum cSub { size 8{k=1} } cSup { size 8{M} } {L rSub { size 8{k} } rSup { size 8{h} } \) ={}} Sum cSub { size 8{k=1} } cSup { size 8{M rSub { size 6{c} } } } { { {ΔL rSub { size 8{k} } rSup { size 8{h} } } over {ΔP rSub { size 8{n} } } } } } {} (14)

Consistent with the conventional DC Load Flow Model, the heating losses in each of the individual circuits can be assessed as follows:

Lkh=rkFk2 size 12{L rSub { size 8{k} } rSup { size 8{h} } =r rSub { size 8{k} } F rSub { size 8{k} } rSup { size 8{2} } } {} (15)

where:

Fk size 12{F rSub { size 8{k} } } {}- per unit active flow through transmission circuit k , and

rk size 12{r rSub { size 8{k} } } {}- per unit resistance of transmission circuit k.

Given that Fk is calculated through the conventional DC Load Flow Model, the Nodal TLF for a particular Node n is now given by

TLFn=k=1MΔLkhΔPn=k=1MΔ(rkFk2)ΔPn size 12{ ital "TLF" rSub { size 8{n} } = Sum cSub { size 8{k=1} } cSup { size 8{M} } { { {ΔL rSub { size 8{k} } rSup { size 8{h} } } over {ΔP rSub { size 8{n} } } } } = Sum cSub { size 8{k=1} } cSup { size 8{M} } { { {Δ \( r rSub { size 8{k} } F rSub { size 8{k} } rSup { size 8{2} } \) } over {ΔP rSub { size 8{n} } } } } } {} (16)

This expression can be further expanded as follows:

TLFn=k=1M2rkFkΔFkΔPn size 12{ ital "TLF" rSub { size 8{n} } = Sum cSub { size 8{k=1} } cSup { size 8{M} } {2r rSub { size 8{k} } F rSub { size 8{k} } { {ΔF rSub { size 8{k} } } over {ΔP rSub { size 8{n} } } } } } {} (17)

The above expression is fundamental for the evaluation of the Nodal TLFs using the required DC Load Flow approach. The sensitivity factor ΔFkΔPn size 12{ { {ΔF rSub { size 8{k} } } over {ΔP rSub { size 8{n} } } } } {} in (17) measures the change in the power flow in circuit k due to an increase in power injection at Node n. In the conventional DC Load Flow Model, these sensitivity factors do not depend on loading conditions but only on the network topology and reactances of the network circuits. Hence, for a network with a fixed topology the sensitivity factors are constant and are evaluated without considering generation and demand.

This is consistent with the requirement set in Section 2.3, point 2(ii). However, the Nodal TLFs (in expression 17) do depend on loading conditions since load flows in individual circuits (Fk) will be driven by loading conditions.

The sensitivity factors, the ratio of the change of power flow Fk, between Nodes a and b, to the increase in power flow Pn at node n can be calculated from the following expression:

(18)

hkn=ΔFkΔPn=ΔΔPn[1xab(θaθb)]=1xab(ΔθaΔPnΔθbΔPn) size 12{h rSub { size 8{ ital "kn"} } = { {ΔF rSub { size 8{k} } } over {ΔP rSub { size 8{n} } } } = { {Δ} over {ΔP rSub { size 8{n} } } } left [ { {1} over {x rSub { size 8{ ital "ab"} } } } \( θ rSub { size 8{a} } - θ rSub { size 8{b} } \) right ]= { {1} over {x rSub { size 8{ ital "ab"} } } } left ( { {Δθ rSub { size 8{a} } } over {ΔP rSub { size 8{n} } } } - { {Δθ rSub { size 8{b} } } over {ΔP rSub { size 8{n} } } } right )} {}

Given that (10) is expressed in the form of

[θ1θ2...θN]=[0......00X22...X2N0,,,......0XN2...XNN][P1P2...PN] size 12{ left [ matrix { θ rSub { size 8{1} } {} ## θ rSub { size 8{2} } {} ## "." "." "." {} ## θ rSub { size 8{N} } } right ]= left [ matrix { 0 {} # "." "." "." {} # "." "." "." {} # 0 {} ## 0 {} # X rSub { size 8{"22"} } {} # "." "." "." {} # X rSub { size 8{2N} } {} ## 0 {} # ",,," {} # "." "." "." {} # "." "." "." {} ## 0 {} # X rSub { size 8{N2} } {} # "." "." "." {} # X rSub { size 8{ ital "NN"} } {} } right ] cdot left [ matrix { P rSub { size 8{1} } {} ## P rSub { size 8{2} } {} ## "." "." "." {} ## P rSub { size 8{N} } } right ]`} {} (19)

the sensitivity factors are obtained by the following expression

hkn=ΔFkΔPn=1xab(XanXbn) size 12{h rSub { size 8{ ital "kn"} } = { {ΔF rSub { size 8{k} } } over {ΔP rSub { size 8{n} } } } = { {1} over {x rSub { size 8{ ital "ab"} } } } \( X rSub { size 8{ ital "an"} } - X rSub { size 8{ ital "bn"} } \) } {} (20)

where:

[000[Yr]1]=[0......00X22...X2N0,,,......0XN2...XNN] size 12{ left [ matrix { 0 {} # 0 {} ## 0 {} # left [Y rSub { size 8{r} } right ] rSup { size 8{ - 1} } {} } right ]= left [ matrix { 0 {} # "." "." "." {} # "." "." "." {} # 0 {} ## 0 {} # X rSub { size 8{"22"} } {} # "." "." "." {} # X rSub { size 8{2N} } {} ## 0 {} # ",,," {} # "." "." "." {} # "." "." "." {} ## 0 {} # X rSub { size 8{N2} } {} # "." "." "." {} # X rSub { size 8{ ital "NN"} } {} } right ]`} {}

Xan size 12{X rSub { size 8{ ital "an"} } } {} and Xbn size 12{X rSub { size 8{ ital "bn"} } } {} are the entries of the inverse of the reduce Ybus matrix (19) positioned in rows a and b, respectively, and in column n.

These factors can be readily computed using matrix techniques routinely employed in load flow calculations. The sensitivity factors only depend on values of network parameters but not on network loading.

The values of the sensitivity coefficients depend on the choice of slack node and therefore, the values of Nodal TLFs (17) will also depend on the choice of slack node. However, the differences in TLFs between any two nodes (TLF differentials) will remain constant irrespective of the choice of slack node, since the differences in sensitivity factors are also independent from the choice of slack node.

The above Nodal TLFs, as defined in (17), represent the incremental change in losses due to an incremental increase in power flow, i.e. incremental generation. Given that the formulas used to calculate TLMOs assume demand oriented definition, the polarity of these Nodal TLFs should be reversed for the subsequent application:

TLFn=k=1M2rkFkΔFkΔPn size 12{ ital "TLF" rSub { size 8{n} } = - Sum cSub { size 8{k=1} } cSup { size 8{M} } {2r rSub { size 8{k} } F rSub { size 8{k} } { {ΔF rSub { size 8{k} } } over {ΔP rSub { size 8{n} } } } } } {} (21)

3 Compliance

The LFM should be compliant with the LFM Specification (this document) at all times. The LFM should not be adopted, nor amendments implemented until the model reviewer has reported on the compliance of the LFM with the specification and the Panel has agreed that the LFM is compliant with the LFM specification.

The Panel is required to agree to any amendment to the LFM Specification, and therefore the LFM, and is required to instruct the TLFA to amend the LFM to comply with the amendments to the specification.

4 Appendix 1 - Definitions and Terms

Adjusted nodal power flows

A form of nodal power flows used to calculate Nodal Transmission Loss Factors in accordance with BSC Section T, Annex T-2.

BSC Year

each successive period of 12 months beginning on 1st April in each year.

Load Periods

Division of the Reference Year into a number of different periods representing typically different levels of load on the Transmission System. Load Periods are mutually exclusive and do not overlap.

Network Data

means the following data relating to the Transmission System:

(i) the identity of each pair of adjacent Nodes;

(ii) for each such pair of Nodes, value of resistance and reactance between the Nodes;

Network data shall be established on the assumption of an 'intact network', that is disregarding any planned or other outage of any part of the Transmission System.

Node

a node is a point on the electrical network at which:

(i) a power flow on to or off the network can occur, or

(ii) two or more circuit (forming part of the network) meet.

A Node refers to nodes on the Transmission System.

Reference Year

12 month period ending 30 September in the preceding BSC Year.

Sample Settlement Period

a representative Settlement Period within a Load Period. Every Sample Settlement Period in the Reference Year falls into one and only one Load Period.

slack node

is a node that acts:

(i) in relation to adjacent nodes, as the reference node for calculating the phase angle of the power flow between nodes.

Transmission Loss Factor

is the factor applied to a BM Unit in a Settlement Period in order to adjust for Transmission Losses.

Transmission Loss Factor Agent

the BSC Agent responsible for producing Zonal Transmission Loss Factors and BM Unit specific Transmission Loss Factors.

Zone

a geographic area in which a GSP Group lies, determined by the Panel but so that the Zones are mutually exclusive and comprise of the whole of (and nothing but) the authorised area under the Transmission Licence.

5 Appendix 2 - Illustrative Example

Consider a simple three-Node network with three circuits in Figure 1, with given metered generation and demand volumes.

Circuit per unit reactances and resistances, assuming 100MVA base, are given in Table 1:

Circuit

Per unit reactance

Per unit resistance

1-2

0.1

0.02

1-3

0.2

0.03885

2-3

0.2

0.04

An advanced (non-standard), loss-inclusive DC Load Flow Model was used to determine individual flows that approximately correspond to the given metered data. This load flow uses a piece wise linear representation of losses in individual circuits. Results of these calculations are shown in Figure 1, with sending and receiving powers in individual lines being presented.

complex image of process

Figure 1: Example system with metered volumes and load flows in individual circuits

Nodal TLFs can be calculated in the process composed of the following steps.

5.1 STEP 1: Adjust metered volumes

From given metered generation and demand data, metered heating losses are found to be 19MW (=233+78-292). Generation and demand are now adjusted to allow the application of the loss-inclusive DC Load Flow Model:

G1=G¯ 1(1L2G¯ )=233(1192311)=225.9 size 12{G rSub { size 8{1} } = {overline {G}} rSub { size 8{1} } \( 1 - { {L} over {2 Sum { {overline {G}} } } } \) ="233" cdot \( 1 - { {"19"} over {2 cdot "311"} } \) ="225" "." 9} {}MW

G2=G¯ 2(1L2G¯ )=78(1192311)=75.6 size 12{G rSub { size 8{2} } = {overline {G}} rSub { size 8{2} } \( 1 - { {L} over {2 Sum { {overline {G}} } } } \) ="78" cdot \( 1 - { {"19"} over {2 cdot "311"} } \) ="75" "." 6} {}MW

D3=D¯ 3(1+L2D¯ )=292(1+192292)=301.5 size 12{D rSub { size 8{3} } = {overline {D}} rSub { size 8{3} } \( 1+ { {L} over {2 Sum { {overline {D}} } } } \) ="292" cdot \( 1+ { {"19"} over {2 cdot "292"} } \) ="301" "." 5} {}MW

Clearly, the total adjusted generation equals the total adjusted demand.

5.2 STEP 2: Calculate network flows consistent using the conventional DC Load Flow Model

From given set of reactances and the network topology, the following Ybus can be formed:

[Ybus]=[15105101555510] size 12{ left [Y rSub { size 8{ ital "bus"} } right ]= left [ matrix { "15" {} # - "10" {} # - 5 {} ## - "10" {} # "15" {} # - 5 {} ## - 5 {} # - 5 {} # "10"{} } right ]} {}

By removing the row and the column from the admittance matrix Ybus that corresponds to the slack node, matrix Yr is obtained. Assuming Node 1 is selected to be a slack, Yr is defined as:

Yr=[155510] size 12{Y rSub { size 8{r} } = left [ matrix { "15" {} # - 5 {} ## - 5 {} # "10"{} } right ]} {}

Given the reduced matrix Yr and power injections, voltage phase angles can be computed:

[0θ2θ3]=[00000.080.0400.040.12][2.2590.7563.015] size 12{ left [ matrix { alignl { stack { 0 {} # θ rSub { size 8{2} } {} } } {} ## θ rSub { size 8{3} } } right ]= left [ matrix { 0 {} # 0 {} # 0 {} ## 0 {} # 0 "." "08" {} # 0 "." "04" {} ## 0 {} # 0 "." "04" {} # 0 "." "12"{} } right ] cdot left [ matrix { 2 "." "259" {} ## 0 "." "756" {} ## - 3 "." "015" } right ]} {}

[0θ2θ3]=[00.060120.3316] size 12{ left [ matrix { 0 {} ## θ rSub { size 8{2} } {} ## θ rSub { size 8{3} } } right ]= left [ matrix { 0 {} ## - 0 "." "06012" {} ## - 0 "." "3316" } right ]} {}

Finally, circuit load flows can be computed:

F12=1x12(θ1θ2)=10.1(0(0.06012))=0.6012pu size 12{F rSub { size 8{1 - 2} } = { {1} over {x rSub { size 8{1 - 2} } } } \( θ rSub { size 8{1} } - θ rSub { size 8{2} } \) = { {1} over {0 "." 1} } \( 0 - \( - 0 "." "06012" \) \) =0 "." "6012" ital "pu"} {} (60.12MW)

F13=1x13(θ1θ3)=10.2(0(0.3316))=1.658pu size 12{F rSub { size 8{1 - 3} } = { {1} over {x rSub { size 8{1 - 3} } } } \( θ rSub { size 8{1} } - θ rSub { size 8{3} } \) = { {1} over {0 "." 2} } \( 0 - \( - 0 "." "3316" \) \) =1 "." "658" ital "pu"} {} (165.8MW)

F23=1x23(θ2θ3)=10.2(0.06012(0.3316))=1.357pu size 12{F rSub { size 8{2 - 3} } = { {1} over {x rSub { size 8{2 - 3} } } } \( θ rSub { size 8{2} } - θ rSub { size 8{3} } \) = { {1} over {0 "." 2} } \( - 0 "." "06012" - \( - 0 "." "3316" \) \) =1 "." "357" ital "pu"} {} (135.7MW)

This is presented in Figure 2.

complex image of process

Figure 2: Nodal Power flows in the example system with adjusted generation and demand volumes (total adjusted generation equal total adjusted demand)

5.3 STEP 3: Determine flow-injections sensitivity factors and compute TLFs

Given the entries of the inverse Yr matrix

[000[Yr]1]=[00000.080.0400.040.12] size 12{ left [ matrix { 0 {} # 0 {} ## 0 {} # left [Y rSub { size 8{r} } right ] rSup { size 8{ - 1} } {} } right ]= left [ matrix { 0 {} # 0 {} # 0 {} ## 0 {} # 0 "." "08" {} # 0 "." "04" {} ## 0 {} # 0 "." "04" {} # 0 "." "12"{} } right ]} {},

flow-injection sensitivities can be computed:

5.3.1 Circuit 1-2

ΔF12ΔP1=1x12(X11X21)=10.1(00)=0 size 12{ { {ΔF rSub { size 8{1 - 2} } } over {ΔP rSub { size 8{1} } } } = { {1} over {x rSub { size 8{1 - 2} } } } \( X rSub { size 8{"11"} } - X rSub { size 8{"21"} } \) = { {1} over {0 "." 1} } \( 0 - 0 \) =0} {}

ΔF12ΔP2=1x12(X12X22)=10.1(00.08)=0.8 size 12{ { {ΔF rSub { size 8{1 - 2} } } over {ΔP rSub { size 8{2} } } } = { {1} over {x rSub { size 8{1 - 2} } } } \( X rSub { size 8{"12"} } - X rSub { size 8{"22"} } \) = { {1} over {0 "." 1} } \( 0 - 0 "." "08" \) = - 0 "." 8} {}

ΔF12ΔP3=1x12(X13X23)=10.1(00.04)=0.4 size 12{ { {ΔF rSub { size 8{1 - 2} } } over {ΔP rSub { size 8{3} } } } = { {1} over {x rSub { size 8{1 - 2} } } } \( X rSub { size 8{"13"} } - X rSub { size 8{"23"} } \) = { {1} over {0 "." 1} } \( 0 - 0 "." "04" \) = - 0 "." 4} {}

5.3.2 Circuit 1-3

ΔF13ΔP1=1x13(X11X31)=10.2(00)=0 size 12{ { {ΔF rSub { size 8{1 - 3} } } over {ΔP rSub { size 8{1} } } } = { {1} over {x rSub { size 8{1 - 3} } } } \( X rSub { size 8{"11"} } - X rSub { size 8{"31"} } \) = { {1} over {0 "." 2} } \( 0 - 0 \) =0} {}

ΔF13ΔP2=1x13(X12X32)=10.2(00.04)=0.2 size 12{ { {ΔF rSub { size 8{1 - 3} } } over {ΔP rSub { size 8{2} } } } = { {1} over {x rSub { size 8{1 - 3} } } } \( X rSub { size 8{"12"} } - X rSub { size 8{"32"} } \) = { {1} over {0 "." 2} } \( 0 - 0 "." "04" \) = - 0 "." 2} {}

ΔF13ΔP3=1x13(X13X33)=10.2(00.12)=0.6 size 12{ { {ΔF rSub { size 8{1 - 3} } } over {ΔP rSub { size 8{3} } } } = { {1} over {x rSub { size 8{1 - 3} } } } \( X rSub { size 8{"13"} } - X rSub { size 8{"33"} } \) = { {1} over {0 "." 2} } \( 0 - 0 "." "12" \) = - 0 "." 6} {}

5.3.3 Circuit 2-3

ΔF23ΔP1=1x23(X21X31)=10.2(00)=0 size 12{ { {ΔF rSub { size 8{2 - 3} } } over {ΔP rSub { size 8{1} } } } = { {1} over {x rSub { size 8{2 - 3} } } } \( X rSub { size 8{"21"} } - X rSub { size 8{"31"} } \) = { {1} over {0 "." 2} } \( 0 - 0 \) =0} {}

ΔF23ΔP2=1x23(X22X32)=10.2(0.080.04)=0.2 size 12{ { {ΔF rSub { size 8{2 - 3} } } over {ΔP rSub { size 8{2} } } } = { {1} over {x rSub { size 8{2 - 3} } } } \( X rSub { size 8{"22"} } - X rSub { size 8{"32"} } \) = { {1} over {0 "." 2} } \( 0 "." "08" - 0 "." "04" \) =0 "." 2} {}

ΔF23ΔP3=1x23(X23X33)=10.2(0.040.12)=0.4 size 12{ { {ΔF rSub { size 8{2 - 3} } } over {ΔP rSub { size 8{3} } } } = { {1} over {x rSub { size 8{2 - 3} } } } \( X rSub { size 8{"23"} } - X rSub { size 8{"33"} } \) = { {1} over {0 "." 2} } \( 0 "." "04" - 0 "." "12" \) = - 0 "." 4} {}

Applying equation (x), nodal TLFs can be calculated:

TLF1=0 size 12{ ital "TLF" rSub { size 8{1} } =0} {}

TLF2=2(r12F12ΔF12ΔP2+r13F13ΔF13ΔP2+r23F23ΔF23ΔP2)=2(0.02x0.6012x(0.8)+0.03885x1.658x(0.2)+0.04x1.357x0.2)=0.0232alignl { stack { size 12{ ital "TLF" rSub { size 8{2} } =2 \( r rSub { size 8{1 - 2} } F rSub { size 8{1 - 2} } { {ΔF rSub { size 8{1 - 2} } } over {ΔP rSub { size 8{2} } } } +r rSub { size 8{1 - 3} } F rSub { size 8{1 - 3} } { {ΔF rSub { size 8{1 - 3} } } over {ΔP rSub { size 8{2} } } } +r rSub { size 8{2 - 3} } F rSub { size 8{2 - 3} } { {ΔF rSub { size 8{2 - 3} } } over {ΔP rSub { size 8{2} } } } \) ={}} {} # =2 \( 0 "." "02"x0 "." "6012"x \( - 0 "." 8 \) +0 "." "03885"x1 "." "658"x \( - 0 "." 2 \) +0 "." "04"x1 "." "357"x0 "." 2 \) = - 0 "." "0232" {} } } {}

TLF3=2(r12F12ΔF12ΔP3+r13F13ΔF13ΔP3+r23F23ΔF23ΔP3)=2(0.02x0.6012x(0.4)+0.03885x1.658x(0.6)+0.04x1.357x(0.4))=0.1303alignl { stack { size 12{ ital "TLF" rSub { size 8{3} } =2 \( r rSub { size 8{1 - 2} } F rSub { size 8{1 - 2} } { {ΔF rSub { size 8{1 - 2} } } over {ΔP rSub { size 8{3} } } } +r rSub { size 8{1 - 3} } F rSub { size 8{1 - 3} } { {ΔF rSub { size 8{1 - 3} } } over {ΔP rSub { size 8{3} } } } +r rSub { size 8{2 - 3} } F rSub { size 8{2 - 3} } { {ΔF rSub { size 8{2 - 3} } } over {ΔP rSub { size 8{3} } } } \) ={}} {} # =2 \( 0 "." "02"x0 "." "6012"x \( - 0 "." 4 \) +0 "." "03885"x1 "." "658"x \( - 0 "." 6 \) +0 "." "04"x1 "." "357"x \( - 0 "." 4 \) \) = - 0 "." "1303" {} } } {}

Table 2: Nodal TLFs for incremental increases in generation and demand

Busbar

TLFs for incremental increase in generation

TLFs for incremental increase in demand

1

0.0

0.0

2

-0.0232

0.0232

3

–0.1303

0.1303

The TLFs that represent the incremental change in losses due to incremental change in demand (21) should be used for subsequent calculations of TLMOs.