Guide to electricity network design and planning – part 2
This is part 2 of a three-part guide aimed at qualified engineers and/or graduates who are beginning their career in an electricity distribution network planning department, or perhaps moving into a planning role after some years in a different role (for example operations, maintenance or construction).
Author(s): Dave Openshaw
The focus of this article is on distribution rather than transmission planning. Nevertheless, the general principles are essentially similar, and since the distinction between the two is becoming increasingly blurred given the proliferation of distributed generation and energy resources, planning engineers new to either discipline should find the three parts of this guide useful.
Part 2 covers legislation and regulation, quality of supply, engineering recommendations and technical reports, investment appraisal, low carbon transition, and innovation and smart grids.
The two complementary parts of this guide are:
- Part 1 – which covers the general principles of network design and planning, security of supply, network design basics, voltage management, load forecasting, electrical losses, and analysis and modelling. It includes a useful 21-point check list which planning engineers may find useful when preparing proposals for network investment; and
- Part 3 – which covers the basics of protection and overall protection schemes.
Legislation and regulation
Electricity networks exist to serve customers, and the broad objectives of how the service is provided, paid for and governed will be laid down in relevant legislation. The task of providing customers with electricity safely and reliably is fairly similar wherever in the world it is done, but the legislative and regulatory environment at a detailed level can be very different. The situation in Great Britain (GB) is described below.
Historical background in GB
The structure of the industry up to 1990 was largely unchanged from 1947 when the pre-war industry was consolidated and nationalised. Prior to 1990 there was little distinction between the functions of parts of the industry, and the companies who run them. The supply of electricity was a regional monopoly undertaken by 14 Area Boards. The term ‘supply’ at that time also encompassed both supply, in the sense of retailing of electricity, and also the distribution and metering of electricity to all consumers.
Nearly all generation of electricity, and bulk transmission at 400 kV, was the preserve of the Central Electricity Generating Board (CEGB). The CEGB had the responsibility to ensure that there was always sufficient generation to meet demand and was responsible for the planning, design, construction and economic operation of all power stations.
The Area Boards and the CEGB were wholly government owned, and managed through a quasi-autonomous body called the Electricity Council. The Electricity Council was not part of the formal apparatus of government, nor were its employees formally civil servants. However, in practice it fulfilled a key function in discharging government policy relating to the industry.
Prior to 1990 the concept of regulation was probably indistinguishable from legislation. The industry was constituted in accordance with various enactments in primary legislation and its operation was constrained by a number of secondary legislation regulations and orders 1 . The Electricity Council’s role included turning political requirements, such as the overall affordability of energy or labour, into practice, e.g. the setting of retail prices. In achieving the desired political ends, the Electricity Council would develop specific policies and practices, many of which became the policies of the newly privatised Area Boards in 1990.
All consumers and prospective consumers of electricity had the right to a supply enshrined in primary legislation. The original legislation dated back to the Electric Lighting Act of 1882, and updated many times thereafter. However, there were no such rights for privately owned generation. The legal structure had no allowance at all for privately owned generation to connect at any place or at any voltage in GB, until as late as 1983, with the Energy Act of that year. This marked the start of the ideological retreat from the state in planning the generation, and consumption, of the country. The act itself had little direct effect, but it paved the way for the development of the liberalisation of 1990.
The distinctions in the pre-1990 industry were between generation, transmission and supply, i.e. between the CEGB and the Area Boards. Because of the integrated nature of the CEGB the generation/transmission boundary was blurred in cost and organisational terms. Within the Area Boards, the exact boundary between supply, metering and distribution was indistinct, with a number of other functions, such as energy efficiency and advice on the industrial uses of electricity also being undertaken.
The 14 Area Boards (12 in England and Wales and two in Scotland) ran the wires. Although the engineering function was distinct, metering and administration were not seen as activities separate from managing customers, setting prices, billing and collecting revenue.
Each Area Board, within a framework set out by the Electricity Council, would negotiate with the CEGB for supplies of electricity in bulk which it would then offer to retail customers via tariffs. All the metering and billing of customers was identified at privatisation as supply functions.
The Generation Division of the CEGB effectively owned and ran nearly all the power generation in England and Wales. In Scotland Generation and Transmission was also integrated into the South of Scotland Electricity Board (SSEB) and the North of Scottish Hydro Electricity Board (NoSHEB).
At privatisation the generation fleet was predominantly coal-fired, with about 20% of the capacity from nuclear. Both coal and nuclear generation comprised a mixture of relatively modern and efficient plant, but with the peaks being met by a number of smaller inefficient coal plant.
The CEGB Transmission Division had planned and built the supergrid in England and Wales, whist that in Scotland was planned and built by SSEB and NoSHEB. By 1990 the supergrid was relatively mature.
In England and Wales, the highest distribution voltage was 132 kV, with the supergrid operating at 275 and 400 kV. In Scotland 132 kV was retained as a transmission voltage; a subtle difference that persists to this day. This is far from a trivial difference – it means that a particular project in England and Wales, connected at 132 kV will be subject to distribution use of system charges, whereas the same project in Scotland would be subject to transmission use of system charges, and different commercial terms and conditions.
Current legal and regulatory arrangements
The Electricity Act 1989 privatised the industry and set the legal and regulatory framework and the structure of the industry. The 1989 Act has been subject to a number of minor modifications over the years, although the fundamentals have not changed.
Generation, transmission, distribution and supply are all separately licensable activities, i.e. it is illegal to undertake any of these activities without a licence. There are a number of specific exceptions; for example, domestic owners of photovoltaic cells are generators (as well as customers) under the meaning of the Act, but there is a class exemption from generation licensing for all generators less than 50 MW in size.
Ofgem administers the licences for industry participants. The licences specify in some detail the tasks and behaviours of licensed companies. In the case of distribution and transmission, the licenses also set out in detail how these companies should charge customers, or other licensees, for their activities. The distribution and transmission licences also set out how the detail of companies should interact with customers wanting to make new connections to their networks, and how the charges should be calculated. It is interesting to note that being connected to networks is a right enshrined in the 1989 Act, but the details of how companies can discharge that obligation is defined in the licences. The licences define the need for technical and commercial codes in the industry, and define how these codes should be governed.
There are currently over a hundred licensed suppliers, and a similar number of licensed generators. There are three onshore transmission licensees and about a dozen offshore transmission licenses. There are 14 original distribution licensees (i.e. existing at privatisation in 1990), and currently eight independent distribution network operators (DNOs). All distribution licensees have the same licensed area, i.e. GB. However, the 14 original licensees have an additional obligation to provide the retail market metering and settlement infrastructure within their original regional footprint. In practice many of the independent DNOs have chosen to provide their own.
Whilst primary legislation provides high level obligations, rights and responsibilities, there is a need for a great deal of practical detail on technical and commercial issues. The broad hierarchy of approach in GB is that under these high level legal obligations, the regulator sets out, in companies’ licences, more detail as to how companies discharge their legal duties, in such a way that companies can easily be held to account for discharging them. In terms of interaction with customers, and to provide a further level of granularity and detail, the licences require the companies to maintain codes. Codes either provide the finest detail that customers need in order to be connected and remain connected, or the codes specifically refer to recognised national or international standards.
The commercial codes are:
- The Balancing and Settlement Code – dealing with the wholesale trading market, and the metering and settlement of all energy trades, wholesale and retail.
- The Connection and Use of System Code – the costs and contractual terms for being connected to, and using, the transmission system.
- The Distribution Connection and Use of System Code – primarily covers the use of DNOs’ systems by suppliers, but also covers some aspects of being connected to a DNO’s system.
- The Master Registration Agreement – The rules for the operation of (legacy) metering systems.
- The Smart Energy Code – for the Smart Metering System.
The technical codes are:
- The Grid Code.
- The Distribution Code.
- The Meter Operator Code of Practice Agreement.
All codes have a governance mechanism, involving a Panel whose role is to keep the code under review and update it. All codes require public consultation on any change, and in many cases, agreement to the change by the regulator. Currently, although the governance processes for each panel are similar, they are far from identical, and at the time of writing there is activity looking into greater harmonisation of governance, including the possible merging of panels or code administrators.
The technical codes exist because the spectrum of standards is not complete or sufficient to deal, at least historically, with all the relevant issues associated with connection to and operation of the network. However, the technical codes recognise the importance and efficiency of international standards and defer to them where appropriate.
Standards tend to concentrate on the specifications of kit, and possibly on its autonomous behaviour. Codes tend to address both equally.
Both codes and technical standards generally evolve slowly; legal, regulatory and governance changes often move more quickly – or at least can change the requirements, leaving codes and standards to catch up.
A further current complication in GB and in Europe is the introduction of European Network Codes. The full effects of this are being progressively implemented. However, although there are significant changes for licensees, code panels and administrators, there should be little observable difference for most users. One area where there will be significant changes is the technical capabilities of smaller generators, with new requirements for frequency management introduced even into domestic scale microgeneration, and fault ride through capability for all generation of 1 MW or greater. At the time of writing, the impact of Brexit on the implementation of European Network Codes is uncertain.
In an ideal world, a new technical need will be most easily articulated in a code, and manufacturers and others will respond and over time, where possible, create a standard that meets the need of the code. This can enable a more detailed and specific need in the code to be scaled back to just a performance requirement, or less. This model implies that Code Panels would have perfect foresight of current and future technical requirements, and the implications of all relevant innovations. Clearly this is not possible so much code activity in general is concerned with reacting to wider developments in the sector.
The above picture becomes more complicated as soon as the technical drivers imply new or changed costs for users. For transmission, there is market structure that allows costs to be debated with users, and mechanisms to charge costs with a certain amount of precision about where they land. For distribution, there is currently no real market structure in GB, so the implementation of new technical requirements that imply new costs can be problematic.
Other primary and secondary legislation
Planning engineers should also be aware of the following Acts and Regulations which are relevant to the planning and design of the electricity system:
- The Town and Country Planning Act 1990. The construction of new overhead lines and substations, other than those which are classed as permitted development (e.g. most distribution substations) will require planning permissions, which in turn may require environmental impact assessments and mitigation measures, and (for overhead lines) landowner consents in the form of wayleaves agreements or deeds of grant. For overhead lines, Section 37 of The Electricity Act requires that an application is made to BEIS. Reconstructions or diversion of existing overhead lines will also require a Section 37 consent, other than as provided for by the relaxations set out under The Overhead Lines (Exemption) (England and Wales) Regulations 2009.
- The New Roads and Street Works Act 1991. This Act stipulates requirements for the installation and accommodation of utilities’ apparatus in public highways, including the maintenance of records showing their locations (e.g. positions of cables in footpaths) and the provision of notices, and coordination, of intended utilities’ works.
- The Street Works (Sharing of Costs of Works) (England) Regulations 2002. These regulations make provision for recharging the costs of any required diversions of electrical apparatus (e.g. for new road schemes) including the treatment of betterment and deferment of renewal.
- The Health and Safety at Work Act 1974. The Act places generic obligations on employers and employees to ensure safe working practices.
- The Electricity at Work Regulations 1989. These regulations set out a range of requirements specific to the design, and working practices associated with, electrical systems.
- The Electricity Safety, Quality and Continuity Regulations 2002. As implied in the title, the Electricity Safety, Quality and Continuity Regulations set out a range of requirements specific to the design and operation of electricity systems, and apply to generators, suppliers, network owners and operators, and meter operators.
Quality of supply
There are a number of ways to measure the key outputs and performance of an electricity supply system. From a customer/user perspective what normally matters is the availability of supply, and its absence from interruptions and similar disturbances. The two most common metrics in use worldwide are the system average interruption frequency Index (SAIFI) and the system average interruption duration index (SAIDI).
SAIFI is the average number of interruptions (usually over one year) experienced by all customers served by the supply system under consideration. It is derived from the total number of customer-interruptions divided by the total number of customers.
SAIDI is the average time that all customers served by the supply system under consideration are off supply (again usually over one year). It is derived from the aggregate time that all customers are off supply divided by the total number of customers.
Customer average interruption duration index and customer average interruption frequency index (CAIDI and CAIFI) are also in use in some jurisdictions. Unlike SAIDI and SAIFI, CAIDI and CAIFI statistics relate only to customers that experience interruptions, rather than the whole of the connected customer base.
CAIDI is the average duration of an interruption experienced by customers that suffer one or more interruptions, and is derived from the aggregate of all customer interruption durations divided by the total number of customer interruptions (it follows that CAIDI can also be derived from SAIDI/SAIFI).
CAIFI is the average number of interruptions experienced by customers that suffer one or more interruptions, and is derived from the total number of customer interruptions divided by the total number of customers affected by faults over the same period (again, usually a year).
In Europe, SAIFI and SAIDI are often referred to as customer-interruptions and customer minutes lost. Historically these were also referred to as security and availability, respectively. For convenience, customer-interruptions are often expressed per 100 customers. This is simply a convention to make the numbers easier to handle and simply divides the actual number of customer-interruptions by 100. For the rest of this section we will concentrate on customer-interruptions per 100 connected customers, which we will abbreviate as CI, and customer minutes lost per connected customer as CML FN2
Another key factor of system performance, due to the use of auto-reclose and auto-switching by distributors, is the number of short duration interruptions. Short duration interruptions are partly defined by how short the interruption is. In GB any interruption under three minutes is classed as a short duration interruption, with three minutes being the nominal time in which any auto-reclose or auto switching should have taken place. This too is sometimes reported as an index: momentary average interruption frequency index, which is the total number of all short duration interruptions divided by the total number of customers served.
In practice there is a deal of complexity to consistently identifying and recording CIs and CMLs. For example, if during a fault restoration process, some customers are restored by switching in, say, 30 min – then, two hours later, as part of further restoration switching, they need to be disconnected for 30 min – should these additional customer interruptions be counted as new interruptions, or have they effectively already been counted as part of the interruptions associated with the original incident? The answer depends on what the statistics will be used for. The overall system performance should include all interruptions to provide a true reflection of reality. However, if the statistics are being used as part of an incentive scheme, it might make sense not to count the re-interruptions in the headline interruption number, although still to count the additional CMLs.
In GB, system performance is incentivised by Ofgem’s Interruption Incentive Scheme, or IIS. This scheme has been running since 2002. In this scheme CIs and CMLs are separately incentivised around targets set by Ofgem. If a network company performs better than the target, the company is rewarded, and conversely if worse than the target, penalised. In such a scheme, the targets and the incentive payments are key parameters and need to be set with great care. The target needs to be in line with customers’ expectations and needs, and ideally should be both stretching and achievable for the network company. It is even more important that the incentive rate is set appropriately. As an incentive scheme is designed to drive company behaviour, the incentive rate must reflect how customers value the behaviour, and yet must be appropriately sized to have a real impact on companies’ behaviour.
Achieving all the above in setting an incentive rate is challenging, not least because arriving at a figure, or set of figures, for how customers value security and continuity of supply is also very challenging. Ofgem have progressively moved the incentive rates they have applied to be nearer to ones that relate to currently accepted values of VOLL (i.e. the value of lost load accepted in overall strategic considerations of operation of the power system).
A practical incentive scheme also needs to cater with events genuinely outside of the network companies’ control (although of course, what these are can be subject to significant debate too), and to have caps and collars on overall incentive effects. The Ofgem scheme in GB recognises events such as storms as exceptional when they inflict more than the number of high voltage faults that would normally occur over 3 days in a single incident. Ofgem also applies caps and collars to the earnings or fines from the scheme, such that a network company’s exposure to the revenue swings from the scheme are limited to typically 2% of the company’s overall revenue.
The exact rules for reporting, and the target setting process, are dynamic, with the rules being improved from time to time by Ofgem in GB. A better understanding of the true complexities of accurate reporting can be found in Annex F (‘Interruptions’) of Ofgem’s Regulatory Instructions and Guidance3 . This is the version issued in 2015 for the GB DNOs’ RIIO ED1 price review period 2015–2023.
An insight into how SAIDI and SAIFI varies across jurisdictions can be found in the Council of European Energy Regulator’s annual report4.
The GB scheme has been perceived as having been very effective. The GB system performance in terms of outages affecting customers has improved markedly since privatisation. The average CI per 100 connected customers (CI) and CML were 100 and 122, respectively, in 1991, and 50.8 and 39.2, respectively, in 2014/15 (excluding exceptional events). Although the most recent figures are not yet available, the strength of the regulatory incentive to improve performance has been carried through into the RIIO5 ED1 price control and the DNOs are currently maintaining similar rates of improvement in these measures.
Fig 1 shows the CI and CML improvement from 1990 levels
Fig 1: CI and CML trends post 1990
In the early years of privatisation improvements were mainly due to peer pressure and a new focus and clarity on customer measures. However, this improvement has been maintained, and the last five years show a continuing strong improvement. In part this is due to DNOs installing significantly more system automation. The costs of the automation are not included in DNOs’ RIIO price control allowances, instead DNOs are encouraged to make these investments based on the incentives they earn from performing better.
The improvements post privatisation are more stark when compared with the decade prior to privatisation (Fig 2).
Fig 2: CI and CML trends pre and post privatisation
The peaks of CMLs include the hurricane of 1987 and the ice storms of 1990/91. It is worth noting that the severe winter storms of 1997 and 1998 were managed considerably better than periods of historic bad weather in the 1980s, with those two years’ exceptional problems barely visible in the post privatisation data. However, a note of caution is still necessary; it was only with the introduction of the IIS scheme in 2002 that an auditable regime of recording was put in place, and any records before that date could be less accurate.
Power quality considerations
Power quality is a generic term used to describe the pureness, quality and consistency of the AC sinusoidal voltage waveform. There are three primary considerations for power quality that the planning engineer needs to be aware of:
- The level of distortion to the voltage waveform as a result of the harmonic content of the current passing through the power system – most simply expressed as ‘total harmonic distortion’.
- The level of voltage fluctuation caused by consumers’ appliances (e.g. due to motor starting currents or welding equipment).
- The degree of voltage unbalance (in technical terms the ratio of the negative and positive sequence components of the voltage).
Planning engineers should be aware that there are specific Engineering Recommendations providing guidance on the assessment of each of the above when considering the potential impact of new loads to be connected to the system; these are:
- ER G5/4-16 – planning levels for harmonic distortion and the connection of non-linear equipment to transmission systems and distribution networks.
- ER P28 – planning limits for voltage fluctuation caused by industrial, commercial and domestic equipment.
- ER P29 – planning limits for voltage unbalance.
It is beyond the scope of this paper to describe these Engineering Recommendations in detail. However, compliance is a requirement under the Distribution Code and so planning engineers should be aware of the circumstances under which the methods of assessment described under these Engineering Recommendations should be applied. Loads (or generation) that might give rise to a need for assessments under the above include:
- ER G5/4-1 – Inverter, rectifier (or converter) connected equipment including: solar PV generation and microgeneration, variable speed motor drives, electric vehicle chargers, motor soft-start mechanisms, and power electronics-based devices generally. At higher voltages, DC traction supplies derived from (say) 132 or 33 kV distribution networks are potentially problematic. Potential mitigation measures include: harmonic filtering, power factor correction, and reducing source impedance (increasing fault level).
- ER P28 – electric motors with high starting currents, electric welding equipment, arc and induction furnaces. A key consideration is the frequency, as well as the severity, of the fluctuations. Again, the higher the fault level (the lower the source impedance) at the point of common coupling with other consumers, the less problematic will be the voltage flicker effect.
- ER P29 – any large unbalanced or single (or two) phase load will give rise to negative phase sequence (NPS) components in the voltage waveform. Typical examples include arc furnaces and AC traction supplies. High levels of NPS voltages (>2% for sustained periods) can give rise to overheating of three-phase motor stator or rotor windings. Potential mitigation measures include rearrangement of phase connections (e.g. where two or more unbalanced loads are closely coupled, rearranging one of the connections might improve overall balance) but, more generally, a connection to a higher voltage level or a point of supply less electrically adjacent to vulnerable appliances (such as AC motors) may be required. The recommendation under ER P29 is that mean background levels of unbalance should not exceed 1%.
Case study of a NPS problem and how it was resolved
A compliant was received from an industrial customer’s site electrical engineer that a number of his AC electric motors were overheating under normal loading conditions. He also reported voltage variations between the three phases of the supply. The factory was supplied at 11 kV from a primary substation on an industrial estate in Hemel Hempstead. An adjacent primary substation served a customer with a recently commissioned single-phase arc furnace through a dedicated 11 kV supply. Both primary substations were supplied from the same 132/33 kV grid substation. The grid substation also provided a 25 kV railway traction supply to the west coast main line via a pair of 132/25 kV single phase transformers.
Tests carried out at the primary substations revealed NPS voltage components in excess of 2%. Subsequent studies of the arrangements serving the arc furnace and the rail traction supply showed that the cumulative effect of the single phase supplies was sufficient to explain the level of the NPS voltage component at 11 kV. However, vector analysis showed that by rearranging the 11 kV supply to the arc furnace (i.e. using an alternative pair of phases) the NPS voltage component should be reduced.
The rearrangement was put in hand and tests confirmed that the NPS voltage component had indeed been reduced to less than 2% and, importantly, the motor overheating problem was resolved.
This example emphasises the need for planning engineers to be aware of the potentially wide system impacts of large imbalanced loads and ensure that appropriate studies are undertaken before finalising network designs and connections points.
Engineering recommendations and technical reports
Having outlined some of the more important Engineering Recommendations relevant to network planning, (ER P2/6, P28, P29, G5/4-1) it might be helpful at this point to make a brief reference to some of the other Engineering Recommendations and Technical Reports that are particularly relevant to network planning. The full catalogue of Engineering Recommendations, Technical Reports, and Technical Standards are available from the Energy Networks Association (ENA) and all Network Operators’ ‘planning department’s’ will have access to these, as should consultancies offering network planning services. At the time of writing, many of these recommendations, standards and reports are under review and so planning engineers should always ensure they have access to the most recent documents. Those marked with an asterisk (*) are formally cited in the Distribution Code and are mandatory for both distribution system users and DNOs.
- ER G5/4-1* – Planning levels for harmonic distortion.
- ER G59/3* – Recommendations for the connection of generating plant to distribution systems.
- ER G74 – Calculation of short circuit currents in three-phase systems.
- ER G81* part 1 – Design and planning of low voltage housing development installations.
- ER G81* part 4 – Design and planning of commercial and industrial underground loads up to 11 kV.
- ER G83/2* – Connection of small scale embedded generators (up to 16 A per phase).
- ER G85 – Innovation good practice guide for energy networks.
- ER G91 – Substation black start resilience.
- ER P2/6* – Security of supply.
- ER P14* – Preferred switchgear ratings.
- ER P17 part 1 – Current rating guide for distribution cables.
- ER P17 part 3 – Ratings for 11 and 33 kV cables with extruded insulation.
- ER P18* – Complexity of 132 kV circuits.
- ER P19 – Planning and design of 132 kV systems.
- ER P24* – AC traction supplies.
- ER P26* – Estimation of prospective short circuit currents for three-phase LV supplies.
- ER P27 – Current rating guide for HV overhead lines.
- ER P28* – Planning limits for voltage fluctuation.
- ER P29* – Planning limits for voltage unbalance.
- ER T8/6 – Cost of losses for transformers used on distribution systems.
Engineering technical reports [note that the following are abbreviated forms of the titles of these documents]
- ETR 115 – Report on the computer program Debut for the Design of LV Radial Networks.
- ETR 116 – Report on voltage unbalance due to AC Traction Supplies.
- ETR 122 – Application of ER G5/4.
- ETR 124 – Actively managing power flows with a single connected Distributed Generator.
- ETR 126 – Actively managing voltage levels with a single connected Distributed Generator.
- ETR 135 – Guidance for the operation and management of fluid filled cables.
- ETR 130* – Application guide for assessing the capacity of networks containing DG (see also ETR 131* – Analysis package).
- ETR 132 – Improving network performance under abnormal weather conditions.
- ETR 136 – Vegetation management near electricity equipment (see also TS 43-8 – Overhead line clearances).
- ETR 138 – Resilience to flooding of Grid and Primary substations.
- ETR 139 – Recommendations for setting of loss of mains protection relays.
It must be emphasised that the above documents are subject to ongoing review and in any case may not always represent ‘latest thinking’ on matters affecting electricity network planning. Nevertheless, with that caveat, they represent a useful baseline reference source of information and guidance both for newly appointed and experienced planning engineers.
As mentioned in the section on General Principles of System Planning in part 1 of the guide, the planning engineer should consider not just the immediate, but also the long-term needs of the network and ensure that, as per the 21-point Planning Engineer’s Checklist in part 1 of the guide, any proposed scheme achieves the objectives in the most efficient way.
There are three fundamental considerations that apply to network investment appraisal: internal rate of return, net present value (NPV) and cost–benefit analysis (CBA) though it is often the case that the techniques are interdependent. For example, it may be necessary to first determine the least cost (in NPV terms) design before calculating the return on capital likely to be achieved, or the potential benefit of any incremental investment that might achieve a higher benefit/cost ratio than the least cost scheme.
Principles of discounted cash flow (DCF)
A fundamental principle of DCF is that a sum of money has a higher value today than in the future (at least with positive or zero inflation). That in turn leads to a discount rate which represents the return that could be made on investment made today. The rate can be expressed in ‘real’ or ‘nominal’ terms; the former does not include inflation and is the rate normally used for network investment appraisal – so one might talk about a discount rate or return of ‘3% real’.
The use of DCF involves adjusting all cash flows, irrespective of when they occur, to today’s values (or the value at any specific year – e.g. the first year of a regulatory review period). It also follows from the ‘time value’ of money must refer to the base year for the appraisal, so one might talk about £(2016). The purpose of the DCF study is to adjust all outward and any inward cash flows to the base year by discounting them (at the discount rate) to their present value. The summation of all future outward and inward cash flows results in a NPV. The following rules apply to DCF studies:
- Non-cash aspects such as depreciation are not included.
- Interest costs are excluded since they are inherently accounted for by the discount rate.
- Income is conventionally shown as a positive cash flow; expenditure as a negative cash flow.
- Precise timing of cash flows is essential.
It should also be noted that the cash flows used in the study will not necessarily be limited to those seen by the network company; for example: it is legitimate to include the cost of losses in the study even though these costs are borne by Suppliers in terms both of the energy cost and use of system charges, and passed on to customers. Moreover, the ‘cost’ of losses might include a factor relating to carbon cost. The inclusion or otherwise of third parties’ costs and benefits will depend on the requirements of the investment case, including who will authorise it.
DCF studies are undertaken using a spreadsheet, cash flows are entered in the appropriate year and the spreadsheet automatically calculates the discount rate to be applied to those cash flows, and hence calculates an overall NPV.
Typical uses of DCF are to compare the NPV cost of alternative schemes. For example, an apparent need for 33 kV network reinforcement due to general load growth might be met in the shorter term by increasing 11 kV transfer capacity between two primary substations, thereby deferring the need for 33 kV reinforcement for several years (depending on the predicted rate of load growth).
Clearly it would be cheaper in pure cash terms to undertake the 33 kV reinforcement immediately, avoiding the cost of the enhanced 11 kV transfer. However, undertaking the 11 kV transfer might be justified because, in NPV terms, the deferred cost of 33 kV reinforcement would be considerably less than the current cost, and the difference between the two greater than the NPV cost of the enhanced 11 kV transfer.
Internal rate of return
The majority of network investment is essentially non-discretionary – i.e. it represents investment necessary to comply with a statutory obligation or licence (or G.Code or D.Code) obligation. However, discretionary investment can sometimes be justified purely in terms of the return it achieves, provided the rate of return exceeds the cost of capital (CoC) raised to fund the investment. A few words on CoC might be helpful at this point.
Cost of capital
There are essentially two elements to CoC: cost of debt and cost of equity. The ratio of the former to the latter is referred to as the gearing ratio. A company with relatively high debt financing would be considered ‘highly geared’. In general terms, cost of debt will be cheaper than cost of equity but it carries a higher risk as the debt has to be serviced (interest paid), whereas dividends on equity (i.e. provided by shareholders) are optional.
Network companies are required to retain an ‘investment grade’ rating by agencies such as Standard and Poor’s and Moody’s, and indeed there is an obligation on the Regulator when determining price review settlements to make it possible for companies to achieve that grade, which means having an acceptable gearing ratio. Network companies will typically access the bond market at appropriate times to secure debt financing for network investment.
Whilst IIR is the general basis of appraising discretionary investment, for regulated network companies, the value of incentives or penalties is generally the primary consideration for undertaking investment that is neither a statutory or licence (or G.Code/D.Code) obligation. Regulation and the various incentive mechanisms are beyond the scope of this document, but a short overview of those affecting investment decisions will be useful to the planning engineer at this point.
Interruption incentive scheme
Is the strongest single incentive for DNOs in terms of the scope for rewards or penalties. As explained in the section on quality of supply, the incentive rate in this case is informed by customer ‘willingness to pay’ surveys conducted on behalf of Ofgem. The survey determines how much more (or less) customers would be willing to pay (through DUoS charges) for a better (or worse) quality of supply performance – i.e. in terms of number and duration of interruptions. Companies that outperform their quality of supply targets (CIs per 100 connected customer and CMLs per connected customer) will be rewarded by the incentive rate – or penalised by that rate if they fail to meet their targets.
Is an overall incentive to limit network investment to that necessary to deliver the ‘outputs’ agreed with Ofgem at the time of the transmission or distribution price settlement. It is complementary to the Information Quality Incentive which encourages companies to propose ‘efficient’ investment plans in their regulatory business plan submissions.
In simple terms, companies that deliver the required outputs (including for example delivering the agreed load and health index profiles in respect of overall network capacity headroom and asset condition) for a lower overall level of investment than their regulatory settlement allows for, they (i.e. their shareholders) will share the savings with their customers (the latter by reducing DUoS charges).
Conversely, if greater investment than allowed for is required to deliver the outputs, then provided it is deemed ‘efficient’ the additional costs will be shared between shareholders and customers (the latter by increasing DUoS charges).
It follows from the above that if discretionary investment to improve quality of supply is calculated to deliver a higher performance per £ invested than the incentive rate, it will be justified in principle. However, the impact of the additional investment on the available Totex incentive that can be earned will also need to be taken into account.
Albeit not currently the subject of a direct regulatory incentive, reducing network losses is a further source of consideration for discretionary investment. In this case, a DCF based cost-benefit assessment should be undertaken to determine whether the NPV of the discretionary investment is positive. The DCF study will take account of all future annual savings in losses resulting from the discretionary investment – i.e. by discounting the future values at the appropriate discount rate. An example of such discretionary investment to save losses would be to specify a larger (or lower loss) transformer and/or a larger cross section cable than the minimum specification required for the scheme.
It will be apparent from the above that, in practice, there might be numerous options for discretionary investment each delivering one or more additional benefits for an additional level of investment beyond the minimum necessary to achieve the primary objective for the scheme. It follows that a mechanism is required to determine the optimum investment level.
Quality of supply investment is a good candidate for incremental CBA since the incremental investment can be directly related to the incremental CI or CML benefits (and incentive rewards). The test benefit ratio for a QoS scheme is obviously the IIS incentive rate – i.e. the justification for a scheme that delivered exactly that benefit ratio would be neutral. A number of optional schemes could then be considered (e.g. undergrounding overhead line sections, using additional remotely controlled section points or automation points, etc.) The cost-benefit of each scheme option would be calculated and compared with the base case. All options that delivered a superior cost-benefit to the base case would be justified but, of those options, the one that delivered the highest benefit to cost ratio for the level of required investment would be optimum scheme.
In practice, any QoS investment would be part of an overall QoS improvement strategy which will have been determined on the basis of generic CBA. Alignment with the overall QoS strategy will therefore be an important consideration in determining the scope and specification for the scheme.
Post investment appraisal (PIA)
PIA is the final stage gate for a network investment scheme wherein the actual and anticipated costs and benefits are compared. This is an important aspect of investment appraisal as it provides valuable learning that should be incorporated within all future investment appraisals. It follows that PIA requires close liaison with all parties that have been instrumental in the planning, design, costing, construction and commissioning of the project. Openness and honesty are essential to effective PIA; if there are significant variances between planned and actual costs, benefits, timescales, project scope, etc. then the root cause of those differences (positive or negative) need to be ascertained so that future investment appraisals can benefit through incorporating the learning points.
In some cases, for example in the case of QoS investment, the benefits will generally need to be appraised over a number of years and will therefore inform periodic reviews of the overall network development strategy.
Low carbon transition
No guidance on network planning would be complete without a brief reference to the impact on network design of a decarbonised energy sector. Many of the established principles on which networks have traditionally been designed are coming under increasing scrutiny due to the increasing impact of decentralised (or embedded) generation, in particular solar PV and wind farms, and the anticipated impact of electric vehicle charging and both domestic and commercial heat pumps.
National Grid’s Future Energy Scenarios referred to earlier in this document paint a clear picture of how the characteristics of both supply and demand could change radically over the period to 2035. National Grid’s System Operability Framework articulates many ‘system’ challenges that will arise as a result of electrification of heat and transport, and as a consequence of displacement of conventional thermal power stations using synchronous generation, with weather-dependent (intermittent) solar PV and wind farms using converter connected generation and doubly-fed induction generators.
- Lower and more variable system strength, impacting fault levels and inertia.
- System balancing challenges arising from potentially rapid ramps rates of wind and especially solar PV generation in variable weather conditions.
- More rapid rates of change of frequency leading to lower system stability.
- Deterioration in power quality – for example deeper and wider voltage depressions widely propagated across the national system as a consequence of transmission faults.
- Increased propensity for harmonic resonance occurring at harmonic lower frequencies.
- Higher peak demands due to electric vehicle charging and heat pump load, most likely to coincide with the current winter evening peak demand period.
A further emerging characteristic of power system architecture is electrical energy storage which has the potential to help resolve some of the system challenges – for example by providing fast or dynamic frequency response to support system stability, and operating reserve to support system balancing. Electrical energy storage will invariably be distribution network connected ranging from ‘grid-scale’ connected at 11 kV or above to domestic energy storage incorporated as part of a customer’s electrical installation supplied at LV.
Optimally located and operated, electrical energy storage is capable of also providing network support (e.g. providing capacity support for an out-of-firm primary substation as an alternative to major reinforcement). Closely coupled with a wind or solar PV farm, electrical energy storage can improve the effective generation capacity factors and hence moderate variances in export levels. Since some GSPs are now exporting from distribution networks to the transmission system and driving both distribution and transmission reinforcements (or constraints), electrical energy storage could relieve that condition by storing a proportion of the generation output until (say) the evening peak demand period, which might also be advantageous to the generator in terms of the market price they realise.
However, if not co-ordinated in terms of its point of connection and operating regime, electrical energy storage may create additional challenges for distribution and transmission systems, exacerbating existing constraints or creating new constraints.
Innovation and smart grids
A subject matter in its own right but of increasing importance in terms of the planning engineer’s ‘toolkit’ is the use of innovative techniques to address the challenges of low carbon transition, and indeed more conventional challenges too.
A wealth of growing knowledge is accessible from reports on LCNF, NIA and NIC projects (as well as earlier IFI projects and Registered Power Zones) which have pioneered and/or developed active network management techniques such as:
- Demand side response.
- Active generation curtailment.
- More sophisticated active voltage control.
- Dynamic plant ratings.
- Application of grid-scale energy storage.
- Power electronics devices (e.g. SVCs, statcoms and ‘soft’ open points) for voltage, power factor and load flow management.
- Distribution phase-shifting transformers (quad boosters) to improve load sharing.
- Fault current limiters to accommodate distributed synchronous generation (e.g. associated with CHP/CCHP systems or waste-to-energy schemes).
- Enhanced automation, protection and telecommunications.
- Contingency management tools.
- Active management of EV charging.
- Enhanced monitoring and state estimation.
- Smart meters and time-of-use or dynamic tariffs to influence load shape.
- Advanced diagnostics to detect condition degradation and incipient faults.
The above listing is by no means exhaustive and planning engineers should regularly access the ENA’s Smarter Networks Portal8 and the DECC/Ofgem Smart Grid Portal9 and other sources of information to keep abreast of innovation in network design and management.
A further important source of information is the Smart Grid Forum website10 which provides access to SGF work stream reports. The WS7 DS2030 report is of particular relevance to distribution planning engineers.
Finally, the IET has published a number of reports on its website11 providing valuable insights into the challenges for, and requirements of, the future power system. The Power Networks Joint Vision report – Handling a Shock to the System – is of particular relevance. This report has led to a further major BEIS sponsored study by the Energy Systems Catapult and the IET into the requirements of the Future Power System Architecture, the reports for which were released in July 2016.
- for example the 1988 Electricity Supply Regulations, replacing the 1937 regulations of the same name
- Remember that in GB CI is per 100 customers, so this formula needs to be ×100
- RIIO – Regulation by Incentives, Innovation and Outputs – Ofgem’s ‘trademark’ for all its energy network price controls. Introduced in 2013.
- Currently under industry review.
- Note that the following are abbreviated forms of the titles of these documents