Western Power Distribution

LOSSES STRATEGY
Losses Strategy

January 2015

For discussion at Stakeholder

Engagement
Version 3 November 2014
 Table of Contents

1 Our Overall Strategy, Objectives and Outputs ........................................... 4

1.1 Strategy ..................................................................................................4
1.2 Objectives ...............................................................................................4
1.3 Outputs ...................................................................................................4
2 Background ................................................................................................ 5
2.1 What are losses? ....................................................................................5
2.2 Technical losses .....................................................................................5
2.2.1 Fixed losses ....................................................................................... 5
2.2.2 Variable losses ................................................................................... 5
2.2.3 Imbalance .......................................................................................... 6
2.2.4 Power factor ....................................................................................... 6
2.3 Theft in conveyance ...............................................................................6
2.4 Which parts of the network produce the most losses? ...........................6
3 The benefits of reducing losses ................................................................. 7
3.1 Societal losses reduction benefits ..........................................................7
3.2 Effect on the WPD Network ....................................................................7
3.3 Cost benefit analysis (CBA) ....................................................................7
3.4 Low Voltage Networks and Distribution Transformers ............................7
3.5 High Voltage (11kV) Networks ...............................................................8
3.6 EHV (33kV and above) Networks ...........................................................9
4 Understanding losses ................................................................................ 10
4.1 Management of Electricity Distribution Losses .......................................10
4.2 Modelling Power Flows and Losses .......................................................10
4.3 Use of Monitoring and Automation .........................................................10
5 Stakeholder input and review ..................................................................... 11
5.1 Stakeholder engagement .......................................................................11
5.2 Losses strategy review ...........................................................................11
6 Present policy ............................................................................................ 12
6.1 Design ....................................................................................................12
6.2 Asset replacement ..................................................................................12
6.3 Ecodesign ...............................................................................................12
6.4 Imbalance and Power Factor ..................................................................12
6.5 Revenue Protection ................................................................................13
6.5.1 Supplier side abstraction .................................................................. 13
6.5.2 Theft in Conveyance ........................................................................ 13
7 Future changes .......................................................................................... 14
7.1 Smart meters and Demand Side Response ...........................................14
7.2 Electrification of heating and transport ...................................................14
7.3 Distributed generation (DG) ....................................................................14
8 Our plans for losses reduction in RIIO-ED1 ............................................... 15
8.1 Improved understanding of losses ..........................................................15
8.2 Transformer sizes ...................................................................................15
8.3 Cable sizes .............................................................................................16

Losses Strategy – January 2015 Page 2

 8.4 Network design .......................................................................................17
8.5 Asset replacement ..................................................................................17
8.6 Voltage Reduction ..................................................................................17
8.7 Demand side management (DSM) .........................................................18
8.8 Network Meshing ....................................................................................18
8.9 HV Imbalance and power factor .............................................................18
8.10 Harmonics ..............................................................................................18
8.11 LV Imbalance .........................................................................................19
8.12 Revenue Protection ................................................................................19
8.12.1 Supplier side abstraction .................................................................. 19
8.12.2 Theft in Conveyance ........................................................................ 19
9 Options for losses reduction beyond RIIO-ED1 ......................................... 21
9.1 Superconductors ....................................................................................21
9.2 Heat Recovery from electricity networks ................................................21
9.3 Energy storage .......................................................................................21
9.4 Active network management for losses ..................................................22
10 Expected outputs in loss reduction through RIIO-ED1 ............................... 23
10.1 Assumptions ...........................................................................................23
10.2 Transformers ..........................................................................................23
10.3 Cables ....................................................................................................23
10.4 Imbalance ...............................................................................................23
10.5 Table of Outputs .....................................................................................24
11 List of Recommendations from the “Management of Electricity Distribution
Network Losses” report. ............................................................................. 25

Losses Strategy – January 2015 Page 3

1 Our Overall Strategy, Objectives and Outputs

1.1 Strategy
Western Power Distribution’s strategy is to ensure that losses from our Distribution
System are as low as reasonably practicable, having regard to the following:

 the Distribution Losses characteristics of new assets to be installed on our

Distribution System;
 whether and when assets that form part of our Distribution System should be
replaced or repaired;
 the way that our Distribution System is currently operated under normal
operating conditions and potential changes in the future; and
 any relevant legislation that may impact on our investment decisions.

1.2 Objectives
The objective of the strategy is to manage losses such that by 2023:

 our networks become increasingly “energy efficient”. That is, system losses
are generally at a defined level of economic efficiency based on the valuation
of avoided energy loss including carbon abatement;
 future investments take losses into account to ensure that the best balance is
achieved between higher cost network investments today to provide lower
costs of energy supplied to customers in the future;
 the overall level of loss managed by WPD is measurable and reportable with
a predicted level of confidence in order for (a) WPD to manage losses and
demonstrate the effectiveness (b) Ofgem to incentivise continued
performance in loss management and (c) other stakeholders to understand
how we are performing; and
 the risk of losses increasing due to the potential for greater demands of
increasing generation and consumption is avoided without undue costs to
today’s customers.

1.3 Outputs
During RIIO-ED1 we plan to undertake:

 the pro-active replacement of 1,996 distribution transformers

 the design intervention for losses on new installation of 8,184 distribution
transformers and 11,880 kilometres of underground cables

These interventions result in a savings of 57GWh of losses through the period. Our
progress through this plan will be published annually as a part of the WPD
Stakeholder Report.

Losses Strategy – January 2015 Page 4

2 Background

2.1 What are losses?

The amount of energy that enters an electricity network is more than the amount that
is delivered to customers. The principal reason for this is that an electricity network
uses energy in the process of delivering power to customers. This is known as a
technical loss.

Another reason for electricity losses is where there is no meter or supplier at the final
connection to record the usage. There are situations where a connection has been
made to our system without authority. The energy used in these connections is not
metered and does not feature in volumes that suppliers register. As a result it is
shown as a loss on our network. This is known as theft in conveyance or illegal
abstraction.

All Distribution Network Operations (DNOs) are obliged to run an efficient and
economic system as a condition of their Distribution Licence. Losses are one
measure of this efficiency. In addition to the requirements of the Licence, reducing
losses is also a key part of the WPD strategy to help us manage our carbon
footprint.

2.2 Technical losses

The total amount of technical loss is made up of a fixed amount (a function of the
network itself, irrespective of the usage of the network) and a variable amount which
is lost depending on the level of load on the network. Variable losses will change as
load alters on the network and is further impacted by the effect of network imbalance
or power factor.

2.2.1 Fixed losses

The fixed element of losses is made up of the energy which is required when
transformers are energised. As transformers require electrically produced magnetic
fields to operate the energy used creating these fields is essentially fixed while they
are switched on.

2.2.2 Variable losses

The variable element of losses is created due to the heating effect of energy passing
through cables and wires. The conductors in our cables and wires all have a small
resistance and they heat up when currents are passed through them. This heating
effect is expressed as a non-linear relationship between the resistance of the
conductors and the square of the current flowing (I2R). The effect in a highly loaded
network, when individual items of equipment are running near or at full capacity is
very much more significant when an item of plant or network is partly loaded.

The resistance of a cable reduces as its cross sectional area increases so the effect
of losses is reduced in larger cable sizes. There is a very similar variable element
created through the wires and windings which are found in all transformers. The
cross sectional area of winding conductors, and the material used to make them,
dictates the level of losses seen in transformers.

Losses Strategy – January 2015 Page 5

 2.2.3 Imbalance
A network which is not balanced across all three phases will have higher currents
than ideal in at least one phase. There will also be currents flowing in the neutral
conductors if they are present. Due to the non-linear relationship of variable losses,
these higher currents can have a significant effect on losses.

Imbalance is found on all parts of the low voltage network due to customers who use
one or two phases having different load consumptions. On the 11,000 volt network
imbalance is due to the uneven distribution of single phase transformers or two wire
spurs. In order to rebalance the network there are physical actions that are required
on the network.

A rural low voltage overhead network could be rebalanced across phases relatively
simply by moving the overhead service connection to a different phase of the
overhead main. This is more difficult on an urban underground low voltage network
which requires existing service joints to be excavated and new joints made to move
customer supplies to different phases.

Interventions to alter connections will help balance customers and load across a
network based on the maximum demands of those customers. It is very complex to
balance load profiles over time, so some imbalance will always occur at certain
times of the day.

2.2.4 Power factor

Another characteristic which will increase losses by increasing currents on the
network is the power factor. Where the power factor is less than unity the current
has to increase to deliver the required amount of power. This has historically been
an issue for installations used by industrial and commercial customers where most
motor loads or power electronic loads were seen. Developments in domestic power
electronics and the adoption of heat pumps means we will start to see this issue
occurring more on our domestic networks.

2.3 Theft in conveyance

The detection of situations where there is no registered supplier at a final connection
point or no meter installed is very difficult. Often detection comes as a result of
investigations for another reason. In many cases theft in conveyance is connected
to other illegal activities, which prompt investigation and detection.

The normal routine of our visits to premises for fault repair and equipment
replacement tasks and the routine of meter operator visits to replace meters will
often expose cases of theft.

2.4 Which parts of the network produce the most losses?

Modelling predicts that the distribution of electricity at low voltage produces the most
losses on our network. The service cable and low voltage cables supplying
electricity to properties account for 25% of our total losses. The distribution
transformers add another 22% with the 11,000 volts (11kV) network accounting for
another 25%. The higher voltages make up the remaining 28%. The percentages
vary between urban and rural networks.

Losses Strategy – January 2015 Page 6

3 The benefits of reducing losses

3.1 Societal losses reduction benefits

Electricity losses for all elements of the electricity network; distribution, generation
and transmission, are included in the settlement processes and form part of a
customer’s electricity bill. Distribution losses currently account for around 7% of an
average domestic customer’s bill. Therefore any initiatives that are taken to reduce
losses will have a positive effect on these bills.

3.2 Effect on the WPD Network

Electricity loss reduction will help us to run an efficient and economic system. In
section 10 we have included a table of savings for the losses interventions that we
have proposed. At the end of the first year of RIIO-ED1 we expect to have saved
7,178 MWh of losses. The savings in losses have a compound effect across the
RIIO-ED1 period and by the end of the period we expect to have saved a total of
57,4204 MWh.

3.3 Cost benefit analysis (CBA)

We have undertaken cost benefit calculations for specific areas of our network. In
section 8 we discuss the implementation of specific actions to help reduce losses
and these are underpinned by CBA calculations. Our CBAs have been calculated
using the Ofgem provided value of losses of £48.42/MWh.

The Ofgem value of losses is the average of wholesale prices over 2011/12,
expressed in 2012/13 prices. It provides a fixed base for losses CBA calculations
and indicates the societal benefits in the reduction of losses. There is no direct
financial benefit to us when using this figure in our CBAs, and the cost of investment
in loss reduction is a direct cost to us.

Only some actions to reduce losses deliver favourable CBAs when considered in
isolation. Actions taken in conjunction with other work tasks on the network are
more likely to show a benefit. For example, whilst it is not beneficial to actively
replace existing underground cables to reduce losses it can be beneficial to oversize
them when they are replaced as a part of other works. We have targeted these
uprating actions in areas where there is a positive CBA.

3.4 Low Voltage Networks and Distribution Transformers

The increasing impact of LCTs such as heat pumps and electric vehicles will have a
greater effect on our low voltage networks. These networks generally operate in a
radial fashion and are sized to accommodate the load seen during normal running
conditions. In this operational configuration the uprating of assets is beneficial and
is unlikely to lead to stranded assets.

CBAs show this and have produced favourable outcomes for the following areas:

 Uprating existing low voltage underground cables in conjunction with other

works

Losses Strategy – January 2015 Page 7

  Uprating existing 11kV/LV ground mounted distribution transformers in
conjunction with other works
 Discontinuing the use of 95mm2 cable for low voltage mains
 Discontinuing the use of 16mm2 cable for services
 Designing LV networks without cable size tapering

The replacement of a ground mounted transformer is a complex task which involves

the replacement or reconnection of associated switchgear and plant. When we plan
a task such as this it is beneficial to uprate the transformer. The replacement of a
pole mounted unit is unlikely to include works on associated switchgear and plant.
CBAs did not produce favourable outcomes for the uprating of pole mounted
11kV/LV transformers in conjunction with other works. We have considered the
range of sizes we use on our network and have shown a favourable CBA for:

 Discontinuing the use of 16kVA (single phase) and 25kVA (three phase) pole
mounted transformers.

Distribution transformers can also be considered for replacement to reduce losses in

a specific area related to the historic loss designs and assumptions used.
Distribution transformers which pre-date 1958 were built to a range of designs and
specifications which preceded the BEBS-T1 specification for distribution
transformers. The BEBS-T1 specification set a maximum level for losses which
harmonised transformer design and set a standard for losses. CBAs show this and
have produced a favourable outcome for the following area:

 Early replacement of pre-1958 ground mounted distribution transformers.

We have further considered the range of ground mounted distribution transformer

sizes we use on our network and have shown a favourable CBA for:

 Discontinuing the use of 315kVA ground mounted distribution transformers.

3.5 High Voltage (11kV) Networks

When we consider the uprating of assets on the 11kV network we do not see the
same levels of LCT impact. The high voltage network cables are generally sized to
accommodate the load seen during normal running conditions and also support
adjacent networks in times of fault. The CBAs for the targeted uprating of HV
underground cables in conjunction with other works do not produce favourable
outcomes. However, the 11kV network will see a general increase in loads as a
result of LCT impacts. We have considered the range of cable sizes we use and, at
11kV, the use of 185mm2 cable in place of 95mm2 cable does provide a losses
benefit.

CBAs show this and have produced favourable outcomes for the following area:

 Discontinuing the use of 95mm2 cable at 11kV

Losses Strategy – January 2015 Page 8

3.6 EHV (33kV and above) Networks
At the higher voltages (33kV and above) there is no benefit in uprating cables in
conjunction with other works. As cables sizes at these voltages are generally larger,
there is not a favourable outcome in specifying a larger minimum cable size. We are
currently retendering our supply contract and do propose to harmonise our 33kV
cable sizes which will provide an associated losses benefit. The CBAs for the
uprating of cables do not produce favourable outcomes.

33kV/11kV Transformers are generally sized and operated in pairs to provide the
ability to supply all customers during a fault. With this configuration it is not possible
to achieve the same level of benefits by replacing these transformers with uprated
units. The CBAs for the uprating of these transformers do not produce favourable
outcomes.

An extrapolation from the HV network can be made onto the EHV and higher voltage
networks. In general terms these networks are sized for normal running and fault
support and do not operate in a radial fashion. The costs also outweigh the benefits
in following area:

 Uprating of EHV transformers (33kV and above)

 Uprating of EHV cables (33kV and above)

Losses Strategy – January 2015 Page 9

4 Understanding losses

4.1 Management of Electricity Distribution Losses

We have recently completed a major study of network losses which builds on
previous research. In conjunction with UKPN we engaged SOHN Associates and
Imperial College London to complete the project which was funded through the
Ofgem Innovation Funding Initiative (IFI). The project sought to progress the
concept of making distribution networks as energy efficient as economically
possible, raising awareness of relevant issues and proposing solutions which take
more account of losses in network design.

The project has created a set of 26 recommendations that are shown in Section 11.
We are already addressing some of the recommendations in this losses strategy and
plan to incorporate more in future iterations of our strategy through RIIO-ED1 and
beyond. Individual recommendations are cross-referenced to our proposals where
appropriate.

4.2 Modelling Power Flows and Losses

‘Carbon Tracing’ is a project that we are undertaking to model power flows and
better understand the effect of different network configurations on losses. Variable
losses will increase with the increased demand seen from LCTs and we are using
the model to target where losses can be reduced. The Carbon Tracing project also
researches the effect of customer engagement and behaviour on reducing losses.

4.3 Use of Monitoring and Automation

Where automation and monitoring is fitted to substations to accommodate the
additional loads brought about by LCTs, these new data and control points can also
be used to help target our plans for loss reduction.

Losses Strategy – January 2015 Page 10

5 Stakeholder input and review

5.1 Stakeholder engagement

Losses reduction activities will be reported annually as part of the WPD Stakeholder
Report. This will set out improvements achieved in the year; actions planned for the
following year and any longer term programmes.

Our losses strategy has been reviewed at a specific Losses Strategy Stakeholder
Event which was held in November 2014. We targeted the event at stakeholders
who have a specific interest and understanding of losses.

As a result of the event we have included new elements in our strategy relating to
service cable selection, harmonic analysis and HV imbalance in our strategy. We
have also included more detail on existing sections relating to DSR, network
meshing, revenue protection activities and the treatment of Ofgem CBAs.

5.2 Losses strategy review

Our losses strategy will be reviewed on an annual basis throughout the RIIO-ED1
period. We expect knowledge to be gained from the range of LCNF projects that will
help us to better understand the profile of customers’ load and the effect this has on
peak demands and peak losses.

Losses Strategy – January 2015 Page 11

6 Present policy

6.1 Design
Our design software takes account of losses when modelling network designs. The
majority of losses on the network occur on the low voltage system, and our
WinDebut LV design programme automatically designs with reference to losses.

6.2 Asset replacement

Our normal programme of asset replacement has an effect on losses. Changing
older transformers for newer models will reduce overall losses as new transformers
have lower losses than old ones.

Where overhead lines are replaced we aim, where possible, to replace small
diameter aluminium conductors of smaller sizes first. The replacement conductors
have a larger cross sectional area and therefore a lower level of variable losses.

6.3 Ecodesign
The EU has implemented a regulation to implement Directive 2009/125/EC
regarding the Ecodesign of electrical equipment which obliges us to install more
efficient transformers. We have recently tendered for the supply of distribution
transformers and have used the new regulations in our specification. At voltages
above 11kV our transformer specifications already exceed the requirements of the
Ecodesign directive. (Recommendation 9)

To comply with the Ecodesign directive manufacturers will have to find ways to
reduce the level of losses inherent in their products. To reduce the variable losses in
a transformer the resistance of the wires needs to be decreased, which can be done
by increasing the cross sectional area of the wires or by using materials with a lower
resistance. To reduce the fixed losses the efficiency of the magnetism needs to be
improved, which can be done by using materials with better magnetic properties.

Improving both of these elements can often result in a transformer with a larger
physical size. As many of the transformers that we use are installed as
replacements for existing units it is not always feasible to accommodate a larger size
due to the space constraints of existing distribution substations and GRP housings.
We have worked with manufacturers to develop more efficient transformers that
retain the same footprint and dimensions as existing units.

6.4 Imbalance and Power Factor

Our standard connection terms follow the format of the National Terms for
Connection clauses. We expect customers to maintain a power factor that is as
near unity as possible and in any case no less than 0.95 lagging. Where a
connection is made across two or more phases customers are expected to manage
their load evenly across the phases provided.

In line with the changes to the Common Distribution Charging Methodology in April
2010, we introduced an excessive reactive power charge for HV and LV half hourly
metered customers.

Losses Strategy – January 2015 Page 12

6.5 Revenue Protection
Our revenue protection activities fall into two main categories, covering the area of
general illegal abstraction where we work with suppliers and the area of theft in
conveyance where electricity is abstracted on the WPD side of the customer’s
installation. We currently provide a fully investigative revenue protection service for
suppliers in all of our licence areas.

6.5.1 Supplier side abstraction

Our investigative work for suppliers uncovers around 6,000 cases per year of illegal
abstraction. Around 1,000 of these cases are related to cannabis production and, as
a result, we work closely with local police. We identify around 2.7 million units per
year which are passed through to suppliers for entry into the settlements process.

We engage with industry partners via the UK Revenue Protection Agency (UKRPA)
and have created a reporting system to make is simple for meter operators to report
incidents of interference to us.

Our work with the police forces has strengthened our relationships with them and the
other emergency services. We now provide assistance on awareness course run at
the East and West Midlands Police and Fire training centres. This training helps the
emergency services detect illegal abstraction and also ensures they remain safe
near installations that have been tampered with.

We also investigate points of connection which are energised but not registered.
Where a connection has become energised without a supplier, it is often very
complicated to unravel the registration process and appoint a supplier. Customers
in this situation have often tried to rectify the situation without success, and we are
able to help them resolve the issue.

6.5.2 Theft in Conveyance

In many cases theft in conveyance occurs when a non-standard connection is made
directly to the WPD network, or where a service to a new property is installed and
connected to our network without our knowledge but using “industry standard”
equipment. It is often difficult to identify these connections, especially where the
property is rented and the landlord states that energy is included in the rental
payments.

We work with housing groups and local authorities in this area, and find that our
police awareness training often helps identify cases.

We have registered a scheme under Schedule 6 of the Utilities Act 2000 which
allows us to take action to recover the monetary value of units abstracted while in
conveyance. We publish our unit price in our statement of charges and have taken
cases to court where appropriate.

Losses Strategy – January 2015 Page 13

7 Future changes

7.1 Smart meters and Demand Side Response

Smart Meters will give suppliers the functionality to offer a greater range of tariffs
including more time of use based tariffs. Where these are used to move load away
from peak times there will be a consequential reduction in overall losses due to the
reduced times when distribution equipment is operating at its maximum rating.

The meters can also be used to schedule certain loads to create a more consistent
and flatter load profile. This will increase the overall level of utilisation of the network
which will reduce the level of physical network reinforcement. However, this load
profile will also increase overall losses by increasing the overall utilisation of our
assets.

The Electricity Networks Association report “Review of Analysis of Network Benefits

from Smart Meter Message Flows” states that the benefit to DNOs of time of use
tariffs and demand side response applications within RIIO-ED1 is £100.8m which
rises to £685.7m within the RIIO-ED2 period. (Recommendation 5)

7.2 Electrification of heating and transport

The Government Carbon Plan places more emphasis on using electricity to provide
energy for heating and transportation. The effect of this is that our network will see a
higher utilisation and, as a result, higher levels of losses. The electrical loads due to
heat pumps and electric vehicles have the potential to be time managed. We may
be able to schedule this demand to help with the control of the total load on our
networks and therefore will have the effect of filling in gaps in our load profiles.
However such technology will also increase the utilisation of our network and
therefore increase losses.

7.3 Distributed generation (DG)

Traditional centralised generation sources are mostly connected to the National Grid
network. This requires energy to be transported through both the transmission and
distribution networks.

Distributed Generation sources are often connected directly to distribution networks.

Whilst this eliminates transmission losses it can lead to higher losses on the
distribution network. This occurs when the generation exceeds local demands and
there is a need to move the energy across the distribution network.

Distributed Generation can also create the opportunities for us. We are investigating
how to use the generator’s output characteristics to manage losses through load
balancing and power factor correction.

Losses Strategy – January 2015 Page 14

8 Our plans for losses reduction in RIIO-ED1

8.1 Improved understanding of losses

Most of the work undertaken on losses to date has been to better understand
network losses rather than the practical options for reducing losses. During the
RIIO-ED1 period we will reach a stage where we can apply this knowledge to ways
that we can actively reduce losses. As we move toward this stage we will take an
iterative approach for our actions. We will highlight those that produce a positive
CBA and develop them.

In order to see the effect of reducing losses we need to be able to set a baseline of
current losses. We plan to do this by using the established highly monitored network
in South Wales which supported the Low Carbon Network Fund LV Templates
project.

The LV Templates project provides us with a monitored network covering a wide

area of South Wales. It can measure the power supplied into this network at HV and
also measure the power delivered from the LV substations. The losses in this
section of network will be due to technical losses, as inaccuracies caused by illegal
abstraction or meter data issues generally occur at LV. We will investigate ways of
extrapolating this data to provide reliable loss baselines for different network types.
(Recommendation 22)

We will also make use of more detailed template data being developed through the
FALCON project to refine our modelling. As smart meter deployment accelerates
the actual data provided can also be used, although aggregation rules may affect the
usefulness of this data. (Recommendation 1)

The LV Templates project has also provided use characteristic detail for specific
substations which we have used to identify areas for future losses focus. One
example of this is the work we plan to address Imbalance, which was identified as a
direct result of the LV Templates data.

8.2 Transformer sizes

The variable losses in a transformer are much lower when the unit is partially loaded
and increase greatly as a unit becomes fully loaded. It is therefore possible to
reduce the overall losses by oversizing transformers when they are installed. Whilst
it is not appropriate to do this in all cases, as not all transformers will become
significantly loaded, there is a case for oversizing transformers in a targeted way.
(Recommendation 8)

Using data from the Centre for Sustainable Energy (CSE) we can forecast that there
will be approximately 7% of our network where the up-take of LCTs would most
likely occur and the investment in oversized transformers can be justified. We
would aim to oversize on average 109 transformers per annum at a cost of around
£0.11m per annum.

Losses Strategy – January 2015 Page 15

Our smallest size of ground mounted transformer is a 315kVA unit. There is benefit
in oversizing these transformers universally on installation. We plan to install a
minimum ground mounted transformer size of 500kVA in RIIO-ED1. We would aim
to oversize on average 448 transformers per annum at a cost of around £0.38m per
annum.

Older designs of ground mounted transformers have much higher losses than new
designs. Whilst it is not efficient to replace all transformers early simply to reduce
losses, we will plan to replace very old units and large capacity units in advance of
their normal asset replacement plan. We will aim to replace 1,996 pre-1958 ground
mounted distribution transformers (our entire fleet of these units) in the RIIO-ED1
period at a cost of around £2m per annum. (Recommendation 10)
Pole mounted transformers are relatively small size and there is little justification in
replacing them to reduce losses. There is benefit in oversizing the smallest of
transformers universally on installation. We plan to install a minimum pole mounted
transformer size of 25kVA (single phase) and 50kVA (three phase) in RIIO-ED1. We
would aim to oversize on average 575 transformers per annum at a cost of around
£0.30m per annum.

8.3 Cable sizes

To reduce the variable losses in a cable the cross sectional area of the conductor
needs to be increased. Once a cable is laid and the ground is reinstated, it becomes
expensive to make alterations to the cable. Our opportunity to reduce losses exists
at the time that the cable is initially installed. The resistance of a 185mm² LV cable
is around half that of its 95mm² equivalent. The additional cost of the cable is less
than £10 per metre which is a marginal cost when compared to the excavation costs
that can be between £50 and £100 per metre. Whilst this cost is marginal, it is only
appropriate to oversize cables in some cases. (Recommendation 8)

For service cables there is no justification in the targeted uprating of cables but we
can demonstrate a benefit in discontinuing our smallest service cable size of 16mm 2.
This will add around £0.33m per year at current costs. When considering the loss
benefits in services we have taken account of the additional effects of non-diversified
load profiles on service cables. (Recommendation 4)

At Low Voltage we will install the next size up for all our cable designs in the RIIO-
ED1 period. This will add around £2.89m per year at current costs.

At 11kV there is no justification in the targeted uprating of cables but we can

demonstrate a benefit in discontinuing our smallest cable size of 95mm2. This will
add around £1.42m per year at current costs.

At 33kV and above there is no justification in uprating cables. We have just renewed
our framework contract for 33kV cable supply and have taken the opportunity to
harmonise our cable sizes with other voltages. We have discontinued the 240mm 2
size and standardised on 300mm2.

Losses Strategy – January 2015 Page 16

8.4 Network design
The general arrangements for network design on the Low Voltage system have
remained unchanged for many years. Our research points to areas where an
amendment to the design policy would create a benefit for losses.
(Recommendation 19)

We will design our LV cable networks without cable size tapering. Networks will be
planned using either 185mm2 or 300mm2 cables. The size chosen for a scheme will
be used throughout that scheme and tapering will not be considered.
(Recommendation 13)

The overall level of losses on the low voltage system can be reduced by minimising
low voltage cable lengths. We will consider the potential of doing this by setting a
maximum feeder length which is shorter than the theoretically possible length based
on simple volt drop and capacity calculations. (Recommendation 12)

Reducing the length of low voltage cables will have the effect of increasing the
density of distribution substations for any single network design. Whilst this will
increase the overall fixed losses on the network, variable losses will be reduced as a
result of the reduced loadings on each transformer. We will compare the loss effect
of different arrangement and we will report our findings in further versions of this
strategy. (Recommendation 11)

We are considering amendments to the way that we calculate transformer capacities

to reduce losses. Current calculations make use of the cyclic rating system which
allows for the transformer to operate above its nameplate rating for short periods.
Whilst this increases overall transformer utilisation, it also increases losses
significantly due to the variable losses increasing in a non-linear way. We will
investigate the effect of introducing a design policy which will reduce our maximum
acceptable design loading to 90% of the nameplate rating for all new transformers
installed. (Recommendation 9)

8.5 Asset replacement

The majority of our network is already established and there is no cost benefit in
replacing it wholesale purely as a method of reducing losses. When we add new
assets to the network or replace existing ones we do have an opportunity to consider
the effect of losses and take them into account. The specific proposals in this
strategy will reduce losses when assets are replaced. For other replacement works
where a more “like for like” approach is taken, improvements in loss levels for newer
equipment will also produce a related saving in losses.

8.6 Voltage Reduction

The LV Templates project provided data on the voltages seen on the low voltage
network. It concluded that there is scope to reduce the network voltage and remain
within the statutory voltage parameters. Reducing the voltage will reduce the overall
demand and makes a contribution to loss reduction.

The voltage on the low voltage network can be reduced in many ways but we have
chosen to change our settings at the primary substation level. At this point, the
voltage change can be made automatically and whilst the network remains
connected.

Losses Strategy – January 2015 Page 17

We have commenced a programme of voltage reduction in our South Wales area
and will monitor the effects using the LV Templates monitors. Based on the results
of monitoring the voltage in the trial area we will consider a programme of reduction
across the other WPD licence areas. (Recommendation 6)

8.7 Demand side management (DSM)

Due to the non-linear nature of variable losses, assets working at their maximum
capacity will lead to significantly more losses than those with a reduced loading. The
scale of variable losses can therefore be reduced by simply reducing the demand on
the network or by reconfiguring networks to transfer loads from highly loaded circuits
to lower loaded circuits.

In our FALCON LCNF project we are investigating the benefits of using commercial
and industrial DSM, with an available reduction of 9MVA, to reduce the peaks on our
network and smooth the load profiles.

8.8 Network Meshing

In our Lincolnshire Low Carbon Hub, FALCON and FlexDGrid LCNF projects we are
demonstrating methods to monitor and automatically reconfigure networks. There
are areas of our network where it is not possible to operate with parallel feeding
arrangements (meshing) due to technical limitations. These can be due to loads and
fault level.

Enabling networks to be meshed will smooth load profiles and reduce the overall
losses. Our LCNF projects have been successful in demonstrating the opportunities
and as the projects progress we will identify areas of our network where the
techniques can be replicated.

8.9 HV Imbalance and power factor

We are developing a project with a solar generation customer that will investigate
the feasibility of addressing imbalance and power factor issues on the 33kV network.
The project will use the customer’s inverter equipment to alter the phase angle of the
generated power. It will also use local storage to set the generated power per phase
to reduce overall network imbalance. The storage can also be used to manage the
overall utilisation of the network.

The Solar Storage project, which will shortly be registered under LCNF Tier1, will
establish how a DNO can interact with a generator to improve overall network
losses.

8.10 Harmonics
Harmonics are generated whenever a non-linear load is connected to our network.
The currents generated by harmonics cause problems on our network and contribute
to increased levels of losses in distribution transformers. Eddy currents in
transformers will increase with the square of the harmonic frequency so can become
significant. Within the UK a derating factor, Factor K, can be applied to transformers
to take account of the harmonic currents.

Losses Strategy – January 2015 Page 18

Our network monitoring does not routinely measure harmonics and assumptions
based on calculations of known harmonics can be inaccurate due to the effects of
diversity, which apply to harmonics in a similar way to network loadings. The only
accurate method of assessing harmonic spectrum is to measure losses using
monitors on the network. We are investigating options to add monitors to the
network.

8.11 LV Imbalance
The monitored network provided through the LV Templates project has highlighted
areas of our network where imbalance is an issue and is increasing losses. We are
developing a project which will investigate and assess methods for correcting
imbalance on the low voltage network. The project will look at the relative costs and
benefits of methods such as the service-by-service rebalancing of customers or the
less granular approach of rebalancing whole sections of the network between joint
positions. It will also research the practicalities of installing three phase service
cables and cutouts in all properties to enable rebalancing to take place at the cutout.
In the project area we will add monitoring along the route of the low voltage cable
and also compare the rebalanced profiles with the previously recorded LV
Templates profiles. (Recommendation 3)

Using LV Templates substations we have identified that imbalance in the low voltage
network can lead to neutral currents of around 35% of the phase current. Using the
societal cost of losses we have calculated that the annual potential savings from
imbalance correction are in the order of £50 to £80 per kilometre of LV network
affected. This means that interventions to balance low voltage networks are likely to
focus on low cost, one off options rather than solutions with an ongoing
management activity.

8.12 Revenue Protection

8.12.1 Supplier side abstraction

We are currently working with partners on an IFI project to establish if it is possible
to fit monitoring equipment at a substation which can detect the presence of heat
lamps used for the cultivation of drugs such as cannabis. The detectors look for the
specific electrical harmonic signature created by the heat lamps. If this project is
successful we will use it to assist with detection and share the findings with other
DNOs.

We completed a trial project in the West Midlands area to investigate connections

which were active but did not have a registered supplier. We focused on one
specific connection type and were successful in discovering connections which were
unregistered and were, as a result, abstracting electricity. We plan to extend this
trial into other geographical and customer areas during RIIO-ED1.

8.12.2 Theft in Conveyance

It is not easy to identify installations which have a connection that is not recorded on
our system without physically visiting premises. To complete a desk top study we
plan to use the “addresspoint” standard which is provided by Ordnance Survey that
applies a unique property reference number to all properties on the mapping
background. Our records hold a reference for all known connections which include

Losses Strategy – January 2015 Page 19

grid reference details. Comparison of these two systems will produce a list of
premises which do not have an electricity supply registered to them. By starting
from the assumption that most premises in the UK have an electricity supply, we can
use this list to establish a subset of premises without a registered supply and which
would require a physical inspection. This desktop analysis will make the targeting of
potential theft in conveyance a much more efficient system.

Losses Strategy – January 2015 Page 20

9 Options for losses reduction beyond RIIO-ED1

9.1 Superconductors
The variable losses in a network are directly related to the resistance of the current
carrying conductors. A superconductor has a very low resistance which significantly
reduces the losses generated. Most superconductor technology, which is available
at present, relies on processes which cool the conductors as a method of reducing
their resistance.

We are aware of developments in Essen, Germany where an oxide ceramic

superconductor has been used to replace a 1km section of a network interconnector
between two major substations. The cooling load for the superconductor must be
taken into account when considering the overall loss reduction of the system. There
are no practical solutions available which can cool conductors in a low voltage
distribution setting, but we will continue to monitor developments in this area.

9.2 Heat Recovery from electricity networks

Transformers generate heat during normal operation. Traditionally this heat is lost to
the atmosphere. Distribution transformers lose heat naturally through convection of
insulation oil and larger transformers can force heat loss through systems of fans and
pumps connected to the convection radiators.

The heat can be used as a commercial product to nearby customers or can simply
be used to heat local substation buildings. We expect research in this area to have
developed through RIIO-ED1 and we will monitor this for a future project. We are
aware that it is not possible to transport the heat over any significant distances so
require users of heat to be located close to substations. We will complete a map
based survey our primary substation sites to assess the opportunities which may be
presented by neighbouring properties that may require heat. (Recommendations 16
& 17)

We are aware of developments in this area in Germany and will investigate the
potential for deployment on a DNO network. If appropriate we will develop trials in
this area. (Recommendation 15 & 18)

9.3 Energy storage

Our LCNF project, BRISTOL, is showing the benefits of energy storage at a
customer’s premises. The project was designed to show the benefits of reducing the
effect of local distributed generation on the LV network by managing the peaks of
generation it creates. A secondary benefit of the energy being stored at the
premises is that there are no losses created through the further use of the LV
network.

Losses Strategy – January 2015 Page 21

9.4 Active network management for losses
One way to reduce the fixed losses on the network is to switch assets off. An asset
on “hot standby” (energised but not actually supplying electricity) will continue to
produce fixed losses. Disconnecting duplicate or reserve assets will reduce losses
but will also affect supply security and therefore has to be carefully considered
before being adopted.

We have analysed the impact of switching off duplicate transformers at a 33kV/11kV

site. We have modelled the effect on customers by using network fault data and
telecoms (SCADA) availability data. Using data from 2013 we have concluded that
the IIS impact of the events where the transfer of customers to a “cold” transformer
far outweighs the losses which are saved as a result. The impact of leaving a
transformer off-line and the impact of the short interruption customer inconvenience
of a successful transfer should also be considered in any assessment.

We will monitor Low Energy Automated Networks, the LCNF Tier 2 project being
undertaken by Scottish and Southern Energy Power Distribution (SSEPD). We
expect this project to provide more information on the balance of risks and
opportunities created by switching off transformers.

Developments in network management systems beyond RIIO-ED1, and the

increased level of monitoring and control will provide a platform for the
reconfiguration of networks to reduce losses without the current concerns over
supply security.

Losses Strategy – January 2015 Page 22

10 Expected outputs in loss reduction through RIIO-ED1

10.1 Assumptions
Whilst it is not practical to accurately measure losses across the whole of our
network we have used normal running and operation conditions to extrapolate the
level of losses that our proposals will achieve

10.2 Transformers
To estimate the savings in our transformer proposals we have calculated the
difference in losses between a transformer operating at full load and the next size up
transformer operating at the same load.

10.3 Cables
To estimate the savings in our cables we have calculated the difference in losses
between a cable operating at full load and the next size up cable operating at the
same load.

10.4 Imbalance
To estimate the savings in imbalance correction we have calculated the losses
created by neutral currents and used this value as the saving. There is also a
reduction in losses due to the reduced loading on a particular phase which is more
complex to estimate, so we have not included this saving in our calculations.

Losses Strategy – January 2015 Page 23

10.5 Table of Outputs

Proposal Interventions Savings per Interventions Savings

per Annum Annum through through RIIO
(kWh) RIIO – ED1 – ED1 (MWh)
Transformers
Replace pre
1958 250 2,694,543 1,996 21,556
transformers
Discontinue
small size pole
575 68,072 4,600 545
mounted
transformers
Discontinue
small size
ground 448 1,140 3,584 9
mounted
transformers
Cables
Discontinue
small size 343 km 412,629 2,744 3,301
service cables
Next size up LV
694 km 3,049,799 5,552 24,398
cables
Discontinue
small size 11kV 448 km 951,421 3,584 7,611
cables
Imbalance
Correct
Imbalance at LV Per substation 1,014 Per substation
substations

Losses Strategy – January 2015 Page 24

11 List of Recommendations from the “Management of
Electricity Distribution Network Losses” report.

Recommendation 1: The network modelling and analysis tools used in the study are
based on calibrated representative network models data. Given the increasing
importance of losses, it would be appropriate that DNOs establish the capability of
modelling and evaluating loss performance of their present and future networks,
under different future development scenarios Under Consideration

Recommendation 2: DNOs to consider carrying out more systematic data gathering

associated with power factor to assess the materiality of the issue and to enhance
the understanding of the costs and benefits of power factor correction at consumers’
premises. The business case for power factor correction may then be developed
Under Consideration

Recommendation 3: Further work is required to assess the extent of the imbalance

problem and to test various solutions, which will not only reduce losses but deliver
many other benefits of a well-balanced network. It may be appropriate to develop
policies and working practices for avoiding excessive imbalance in future Under
Consideration

Recommendation 4: The inaccuracy of loss calculation using half-hourly data at the

edges of the LV network should be recognised when conducting network studies
Future Consideration

Recommendation 5: As the benefits of peak demand reduction may be material an

assessment of the opportunities enabled by alternative smartgrid techniques to
achieve this should be carried out Adopted

Recommendation 6: As the benefits of active voltage control in LV distribution

network may be significant, comprehensive assessment of the opportunities to
further reduce network losses should be carried out Under Consideration

Recommendation 7: When considering active network management solutions and

technologies to facilitate low-carbon connections, the impact on losses should be
given full consideration Future Consideration

Recommendation 8: There is a clear case for fundamentally reviewing cable and

overhead line ratings to ensure that future loss costing has been included in the
economic rating calculation. This could be based on Ofgem’s loss investment
guidelines or on loss-inclusive network design standards Adopted

Recommendation 9: The transformer loss calculations indicate that the benefits of

investing in low-loss transformers may be significant and this should be considered
further to establish or otherwise the low-loss transformer business case in line with
UK energy and carbon policy Under Consideration

Losses Strategy – January 2015 Page 25

Recommendation 10: In future, losses may drive early asset replacement when
economically efficient. If early replacement programmes are economically justified
and capable of being funded, appropriate resources would need to be made
available to facilitate delivery of such programmes Adopted

Recommendation 11: Network designers may consider the option of installing

additional distribution transformers to minimise LV network reinforcement cost and
reduce losses Under Consideration

Recommendation 12: In the light of future developments, particularly in relation to

the integration of low carbon demand and generation technologies, it may be
appropriate to reconsider long-term distribution network design. This may take a
strategic view of future voltage levels and include consideration of losses in the
decision-making. Under Consideration

Recommendation 13: In order to reduce losses and provide future flexibility within
LV networks, LV tapering policy may be re-examined. Adopted

Recommendation 14: A review of DNOs’ network modelling and analysis tools and
capabilities may be required to support design engineers in applying new policies
and processes relating to loss-inclusive network design. Future Consideration

Recommendation 15: There is opportunity for considerable further learning in

Europe and also from National Grid. It would be beneficial to share experiences of
waste heat recovery installations among DNOs. Future Consideration

Recommendation 16: An Innovation Project, based upon learning from this initial
Study, may be initiated in order to gather further insight into the technical and
practical solutions which can be tested at more sites. The Project could be scoped to
also tackle the regulatory and commercial market structural issues which will also
need to be overcome to bring heat recovery and use into mainstream application
Future Consideration

Recommendation 17: DNOs may maintain an awareness of the potential for heat
recovery when planning the installation of EHV transformers and seek to install more
systems where the recovered heat may be of commercial use. Future
Consideration

Recommendation 18: Further work on heat storage may be integrated with future
trials work on recovery of heat from the distribution network, as it may improve the
economics of more basic heat recovery systems. Future Consideration

Recommendation 19: DNOs should develop loss-inclusive network design

strategies, based on their specific data, in order to ensure that the overall economic
network operation and design criteria are met. This should include network
modelling capability for answering “what-if” questions in order to predict the impact
of proposed network polices, projects and network demand forecasts on the overall
reported network losses. Under Consideration

Losses Strategy – January 2015 Page 26

Recommendation 20: DNOs, with support from DECC and Ofgem, may determine
the common basis in relation to loss mitigation and loss-inclusive network design
and investment. Future Consideration

Recommendation 21: There is a need to establish the basis for assumptions on

future electricity costs and carbon prices that would be used in loss-inclusive
network investment that is consistent with the overall UK low carbon policy. Future
Consideration

Recommendation 22: Early in the RIIO-ED1 period, DNOs may develop more
accurate means of measuring and reporting of distribution network losses. Under
Consideration

Recommendation 23: The DECC/Ofgem comparison of reported losses shows a

discrepancy which may cause a distorted view of GB DNO losses, within industry,
government and internationally. Future Consideration

Recommendation 24: DNOs may grasp opportunities as they may arise to influence
loss reporting in other countries and as it is presented in international studies. This is
in order to ensure that GB DNOs’ loss management performance is presented
accurately. Future Consideration

Recommendation 25: Industry, government and regulators should consider

developing appropriate regulatory and commercial frameworks that would facilitate
development of loss-generated heat schemes where economically justified. Future
Consideration

Recommendation 26: DNOs’ loss strategies may be “stress tested” to demonstrate

that they can deliver an objective of achieving an economic level of losses based
upon avoided loss valuation, engineering costs and future network demands.
Adopted

Losses Strategy – January 2015 Page 27

 Western Power Distribution (East Midlands) plc, No2366923
Western Power Distribution (West Midlands) plc, No3600574
Western Power Distribution (South West) plc, No2366894
Western Power Distribution (South Wales) plc No2366985
Registered in England and Wales
Registered Office: Avonbank, Feeder Road, Bristol BS2 0TB

www.westernpower.co.uk @wpduk
Losses Strategy – January 2015 Page 28