services report a supporting document to … · services report a supporting document to our...

SERVICES REPORTA SUPPORTING DOCUMENT TO OUR INTEGRATED TRANSMISSION PLAN 2017SEPTEMBER 2017

SERVICESREPORT2017

II

C O N T E N T S

CONTENTS ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I I

INTRODUCTION ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

PART 1: HOW OUR SERVICES FRAMEWORK RELATES TO THE SERVICE PERFORMANCE MEASURES ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

PART 2: OUR CURRENT RCP2 SERVICE PERFORMANCE MEASURES ... . . . . . . . . . . . . . . . 8

PART 3: STAKEHOLDER ENGAGEMENT AND THE REFRESH OF MEASURES FOR RCP3 ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

PART 4: HOW WE USE THE TARGETS AND FUTURE DEVELOPMENT ... . . . . . . . . . . . . . . . 30

APPENDIX A: TYPES OF GRID OUTPUT MEASURES IN THE CAPEX IM ... . . . . . . . . . . . . . 33

APPENDIX B: OUR CURRENT PERFORMANCE AGAINST THE SERVICE PERFORMANCE MEASURES ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

SERVICESREPORT2017

1

I N T R O D U C T I O N

This report describes service performance measures for our regulated transmission business. It explains our services framework, how we apply the framework and the measures that are in place in our current regulatory control period, RCP2.

We are currently in the process of refreshing and evolving our service measures and targets for the next regulatory control period, RCP 3 (2020-2025). This year we have also included the results from our stakeholder engagement process which will be a key input to defining proposed regulated service measures for RCP3.

During RCP2 we have undertaken an initiative to develop our asset health models, data and processes. This initiative will allow us to improve our asset health reporting for the remainder of RCP2. We intend to develop revenue linked asset health output measures for RCP3 using this updated reporting methodology (replacing our current volumetric asset health targets).

We prepare an Integrated Transmission Plan (ITP) every five years, and provide updates most years.1 The ITP provides an overview of our 15 year plans for our regulated transmission business, and consists of:

• The Integrated Transmission Plan Narrative (ITP Narrative), which provides an overview of our plans, an introduction and integrated view of our business, and a roadmap to the more detailed information provided in the supporting documents.

• The Asset Management Plan (AMP) which documents the management processes and outcomes of our asset class lifecycle management for our grid, ICT and business enabling assets.

• The Transmission Planning report (TPR), which describes the management approach and enhancement and development investment requirements for the grid.

• This Services Report, which sets out proposed Grid Output Measures and target performance levels.

1 This satisfies requirements in the Transpower Capital Expenditure Input Methodologies (Capex IM) published by the Commerce Commission to prepare an ITP every five years, and to provide updates most years. The Capex IM is available on the Commerce Commission’s website. This document fulfils the requirements of the Outputs and Performance Objectives Report required by the Capex IM as a supporting document for the ITP.

SERVICESREPORT2017

2

The 2017 ITP describes our regulated transmission business for the 15-year period from July 2017 to June 20322 (the forecast period). Figure 1 illustrates the ITP document structure.

Figure 1: The 2017 ITP Narrative and support ing documents

The 2017 ITP is the third update since our 2013 ITP, which was embedded within our regulatory proposal for RCP2.

The Services Report updates our 2015 Service Measures Report which introduced a new services framework for our business that encompasses both our regulatory measures and other services that we provide (such as customer-funded new connections).

STRUCTUREOFTHISDOCUMENTThis Services Report has four main parts. Parts 1 and 2 provide an overview of our services framework and our RCP2 service performance measures and targets. Part 3 provides an overview of our stakeholder engagement and our proposed service measures for RCP3. Part 4 discusses how we currently use the targets, the changing state of our industry and how we plan to further develop the measures and targets.

Our performance against the existing measures can be found in Appendix B.

2 For readability, unless otherwise stated, the dates utilised in this document represent the financial year ending. For example, the financial year spanning July 2017 to June 2018 financial year is simply referred to as 2018.

AssetManagementPlan(AMP)

ITPSchedules

ITPNarrative

The2017ITP

TransmissionPlanningReport(TPR)

ServicesReport

Supp

ortin

gDo

cuments

SERVICESREPORT2017

3

P A R T 1 : H O W O U R S E R V I C E S F R A M E W O R K R E L A T E S T O T H E S E R V I C E P E R F O R M A N C E M E A S U R E S

In this Part we:

• summarise our services framework • describe how the services framework relates to our service performance measures • introduce our Asset Health Measures

Our services framework identifies eight services, including our system operator business and other activities not regulated under our price-quality path. The services are summarised in Figure 2 below. The first three of these are the most significant drivers of cost and quality for end consumers and are the focus of this report.

Figure 2: Our Services Framework

The remainder of this section sets out the context for our services framework and describes, for our three main transmission services, how they relate to our RCP2 service performance measures.

• Keep interruptions to a very low level, and restore supply quickly when there is an interruption Reliable grid

• Keep sufficient grid capacity and resilience available to allow New Zealand's lowest cost sources of supply to be used to meet demand Grid availability

• Communicate with our customers when supply is interrupted so we can achieve the best outcomes for end consumers Event communications

• Work with customers to connect their assets to the grid, and plan and deliver changes to their connections Grid access

• Safely host customer equipment on our sites Site access

• Provide planning and other information to assist connected parties to make informed decisions Information provision

• Assist in the identification, selection and execution of options to relocate transmission infrastructure Asset relocation

• Operate a competitive electricity market and deliver a secure power supply System operation

SERVICESREPORT2017

4

CONTEXTUnder the Capex IM, we are subject to a ‘quality incentive mechanism’. This requires us to measure and report on our quality of service and includes a mechanism to link a portion of our revenue to reported performance. We will include a proposed incentive regime - based on some of our measures - in our December 2018 proposal for RCP3.3 The proposed regime will include two ‘Grid Output Measures’ (as defined in the Capex IM) - Reliable Grid (measure of grid performance) and Grid Availability (measure of asset performance). We have included several additional measures in this Services Report - these are not Grid Output Measures, but do reflect aspects of our service that are important to our customers. Appendix A describes the different types of Grid Output Measures defined in the Capex IM, and describes how our RCP2 measures relate to each. Appendix B describes our performance against the current service performance measures.

RELIABLEGRIDThe level of reliability experienced by our direct customers is heavily influenced by past investment decisions about the assets employed — their configuration, capability and quality — and how they are maintained. The historic choices involved trade-offs between cost and service level. The choices we make now regarding operation, maintenance, replacement and enhancement (new assets) allow us to incrementally refine performance and cost over time. Prior to RCP2 we measured reliability using ‘system minutes’. This provides a system-wide measure of how much energy demand is unmet due to unplanned interruptions. Targets were set based on sustaining historic performance levels. This approach supports comparison between transmission businesses but does not provide good information on the experience of individual customers. For RCP2 we developed a new approach that measures how often a customer experiences an unplanned interruption, and how quickly we restore supply (both average restoration times, and restoration times that we achieve nine times out of ten). We defined five different service categories (high priority, important, standard, generator and N-security) and set forward-looking targets for each category. This approach means we have 15 reliability service targets (known as Grid Performance Measures) for RCP2: GP 1,2, and 3 for each of the five service categories. These are described in more detail in Part 2.

GRIDAVAILABILITYWe do not make the full capacity of the grid available at all times because we need to take equipment out of service for maintenance and upgrade work, and because equipment sometimes fails. In the vast majority of cases, making assets unavailable does not interrupt supply to customers or end consumers but it can:

• influence reliability, through increasing the risk of unplanned interruption

3 While we propose an approach, the Commerce Commission determines the incentive regime to be applied during RCP3, including the Grid Output Measures and targets.

SERVICESREPORT2017

5

• reduce the efficiency of the electricity supply chain. Increased loading on in-service assets increases losses. In addition, when the capacity of the grid is reduced and constraints occur or are introduced it may not be possible to access the lowest cost mix of generation. This can increase energy costs, and can also cause price separation between pricing nodes4, which increases trading risk (and hence cost) for market participants.

In addition, production and trading decisions made by market participants are affected by our outages, so the stability and predictability of our outage planning is important.

In RCP1 we measured and reported on the availability of our inter-island high-voltage direct current (HVDC) link and the availability of our high-voltage alternating current (HVAC) circuits. We refined the HVAC measure in RCP2 by capturing only circuits that have the most impact on losses or constraints. For both RCP2 measures, targets were set based on forward-looking analysis of achievable performance. For RCP2, these availability service measures are known as our Grid Availability Measures - AP1 and AP2. These are described in more detail in Part 2.

EVENTCOMMUNICATIONSEffective communication during an unplanned interruption helps our customers deal with the impact on their business — direct customers can make better production decisions, distributors can better manage their network and communication with retailers and consumers, and generators can better manage their resources and their trading position.

We did not measure performance against an event communications measure in RCP1 and did not define formal measures for RCP2. However, during RCP2 we have been trialling some performance development measures, including three that focus around time to make initial contact, time to communicate expected restoration time and accuracy of that information. More details concerning our trial measures can be found at the end of Part 3.

Since the RCP2 submission, in acknowledgement of the importance of event communications, , we have also developed a new customer post-event survey. This targets our customer service and communication during planned and unplanned outages, as well as during project works.

ASSETHEALTH

We use asset health data and models to understand the condition of our grid assets and the probability of these assets failing today and in the future. They allow us to plan to address potential problems caused by assets near the end of their useful life through asset refurbishment and replacement. Asset Health is one of several key inputs into our new decision-making processes, aimed at understanding and managing the current and future risk profile of the grid.

4 The New Zealand electricity market sets prices at more than 200 pricing nodes across the transmission grid. Prices are set at a level that reflects the marginal cost of supply at each node. If grid capacity is constrained, then prices downstream of the constraint increase relative to prices upstream of the constraint. This creates a financial risk for retailers and other market participants.

SERVICESREPORT2017

6

We support the use of asset health for reporting purposes and as well as the inclusion of asset health output measures within the incentive regime. Asset health reporting provides our stakeholders, including the regulator with a view of the state of our assets, highlighting potential work required to improve the condition of the grid. An emphasis on incentives is particularly important given the increasing need for us to be flexible and agile in responding to our strategic environment.

In RCP2, we were set several annual and periodic (five year) asset health grid output measures that were linked to revenue. These measures cover six portfolios in our replacement and refurbishment programme chosen because we had existing health data in these portfolios and they are large portfolios with relatively uniform outputs that are discrete and identifiable. Three asset classes were given annual output targets, while an additional three asset classes have targets to be met by the end of the control period5.

Our concerns with the current Asset Health Grid Output measures

The asset health grid output measures were designed to incentivise Transpower to deliver its asset replacement and refurbishment programme. The incentive rate was created to match any incentive benefit received by Transpower for the capex expenditure incentive6, where any underspend was due to non-delivery of planned asset replacement.

The current asset health output targets are therefore not true asset health measures but are primarily volume (works delivery) targets. They assume that Asset Health is the only relevant input into determining asset replacement or refurbishment. This is often not the case as planners will consider a number of factors for investment including maximising efficiency of delivery and minimising impact to consumers when deciding to replace or refurbish an asset. The relevance of asset health to decision making also varies depending on the type of asset.

The current targets and output framework also do not account for varied changes in the condition of assets over a control period. For example, we may collect updated condition data of our conductor fleet that signals it is in better shape than expected. We would then update our models resulting in a ‘healthier’ score for those assets. Based on this information we would decide to defer asset replacement resulting in less expenditure and therefore savings for our customers. However, the current regime would penalise us for not replacing the asset, even though the condition was better than when the investment plan was created.

There are additional scenarios where we would want to divert from a prescribed investment plan. Other reasons include the identification of more innovative or cost effective solutions that produce the same or superior outcome but do not use the same solution (asset replacement) that was originally planned. We therefore see merit in alternative output measures that encourage us to find innovative, lower cost solutions, and allows us to be flexible in response to changing circumstances by to incentivising outcomes and not outputs. As we discuss further in chapter 3, we are in the process of developing a more sophisticated set of asset health grid output measures for use in RCP3.

Grid output measures that are not revenue-l inked

5 The main reason for this differentiation was due to the size of the population of the asset classes. 6 The base capex expenditure adjustment is a symmetric incentive that applies to underspend and overspend

across the base capex allowance.

SERVICESREPORT2017

7

In addition to the revenue linked asset health output measures, the current IPP requires us to report against three additional grid output measures not linked to revenue. These grid output measures have targets relating to the average remaining life (in years) of Transpower’s transmission tower coating, transformers, and outdoor circuit breakers and have associated reporting requirements.

These measures form part of a pilot reporting on asset health measures to ensure the revenue-linked asset health measures are providing appropriate incentives and with a view to linking these measures to revenue for RCP3. We obtained an exemption from reporting on these measures in 2016 so that we can develop improved measures that align with our evolving asset health models and provide more meaningful information for stakeholders.

SERVICESREPORT2017

8

P A R T 2 : O U R C U R R E N T R C P 2 S E R V I C E P E R F O R M A N C E M E A S U R E S

In this part we provide an overview of our current RCP2 service performance measures (introduced in Part 1), how they are applied and targets that were established.

Our service performance measures for RCP2 fall into two categories:

• revenue-linked grid output measures

• non-revenue linked development measures.

Revenue-linked Grid Output measures For RCP2 the Commerce Commission determined 23 revenue-linked grid output measures, grouped into three areas:

• Fifteen rel iable grid measures

• Two grid avai labi l i ty measures

• Six asset health (works delivery) measures.

A target, cap, collar, and incentive rate is defined for each of the revenue-linked measures. The cap and collar set the range of performance within which we are penalised or rewarded. The caps and collars are symmetric, meaning that the incentive rate is the same for rewards and penalties, and the maximum reward is equal to the maximum penalty. If we exactly meet the target, then we do not receive any reward or penalty. The strength of each financial incentive is determined by the incentive rate. The rate is applied to the difference between actual and target performance, within the defined limits. Each measurement area is described further below.

SERVICESREPORT2017

9

a. Reliable grid measures: Three measures for grid reliability were set. These are the number of unplanned interruptions each year, the average duration of unplanned interruptions each year (in minutes), and the 90th percentile duration of unplanned interruptions each year (in minutes). Customer feedback informed us that the frequency and duration of interruptions is important, but expectations as to the level of performance varied between customers. For example, reliability expectations are higher for a large central business district than for most rural areas. Accordingly, sites were categorised into five service categories - high-priority, important, standard, generator or ‘N-security’7 - to reflect these differing expectations, resulting in a total of 15 measures.

b. Grid availabi l i ty measures: These are concerned with the impact that grid asset availability can have on access to the lowest-cost mix of generation. The two grid availability measures set for RCP2 are the availability of our inter-island high-voltage direct current (HVDC) system and the availability of selected high-voltage alternating current circuits (HVAC). The HVAC availability measure is the percentage of time that a selected set of 220 kV circuits is available during a year. For RCP2, we selected 27 circuits based on the potential effect on the electricity market of an outage of that circuit.

c. Asset Health measures: There are works delivery measures for six of our asset classes. The measures are the number of towers painted, grillages commissioned, insulators commissioned, outdoor circuit breakers commissioned, transformers commissioned, and outdoor to indoor conversions commissioned. For RCP2, these measures have volumetric output targets.

Non-Revenue-linked development measures In addition to the revenue-linked grid output measures, the Commission also established a set of suggested ‘performance measure development measures’ for RCP2. These would be tracked for information purposes, with a view to possibly including them as revenue-linked measures in later regulatory periods.

The selected performance measure development measures were identified through consultation with stakeholders.

7 N-Security is when a connection is served by a single line or a single transformer and so any fault on that connection leads to a power outage. Most of our customers have N-1 security, which means they will only experience an interruption at their point of service if there are concurrent equipment outages. This can happen when there are multiple equipment failures, or a single equipment failure at a time when other equipment is out of service for maintenance.

SERVICESREPORT2017

10

RELIABLEGRIDMEASURESReliable grid measures and targets are concerned with the most important aspect of grid service performance for most customers — keeping interruptions to a very low level, and restoring supply quickly when there is an unplanned interruption.

RCP2 Point of service categorisation A key feature of our reliability measures is that we have categorised each point of service into service categories. This recognises that customer expectations and interruption impacts vary across the country, so we should not have a one-size-fits-all approach. For RCP2 we assigned all points of service to one of five categories. Table 1 outlines our categories with a brief description and some examples.

Table 1: Point-of-service categories

CATEGORY DESCRIPTION EXAMPLES

HIGH PRIORITY Very large or essential direct customers

• Bream Bay (oil refinery)

• Hobson Street (Auckland CBD)

IMPORTANT Key industrial direct customers, or large number of end users

• Kaiwharawhara (Wellington suburbs)

• Motonui (methanol plant)

STANDARD

Balance of direct customers

• New Plymouth

• Glenbrook (steel mill)

• Southdown (rail)

GENERATOR Grid-connected generators

• Roxbrough (hydro generator)

• Huntly (thermal generator)

N-SECURITY Customers served by a single line or transformer

• Ohakune

• Poihipi (geothermal generator)

Our 2013 ITP provides a more detailed explanation of our approach to assigning point of service categories for RCP2.8

RCP2 Reliability measures We record every unplanned interruption and its duration. This data informs three reliability measures:

• the number of unplanned interruptions at each point of service (GP1) • average duration of unplanned interruptions at each point of service (GP2) • duration of long unplanned interruptions (GP3). ‘Long unplanned interruptions’ are those with

a duration exceeding 90 per cent of unplanned interruptions (P90).

The use of two types of duration measure recognises that we should aim to reduce the duration of all unplanned interruptions and also maintain an emphasis on the small number of very long unplanned interruptions.

8 Service Performance Measures Report, 2013

SERVICESREPORT2017

11

Very short (less than one minute) interruptions and interruptions relating to operation of New Zealand’s automatic under-frequency load shedding (AUFLS) system are excluded. All other unplanned interruptions caused by faults on our network are included9. Extreme events (e.g. severe storms or earthquakes) are not excluded but the maximum duration attributable to any event is capped at seven days.

Full definitions of the RCP2 measures are available in our RCP2 individual price-quality path determination.10

RCP2 Reliability targets In developing the RCP2 targets we considered what performance we should be able to deliver given the makeup of the grid. We cross-checked them against our estimated Value of Lost Load (VoLL). Generally the RCP 2 targets determined using these considerations represent an improvement on our historic performance.

To assess overall performance we aggregated the RCP2 targets across the different point of service categories. This provided a view of overall historic performance and set targets for the regulatory period.

GP 1: long-term number of unplanned interruptions performance targets Most customers have N-1 security, meaning they will only experience an interruption at their point of service if there are concurrent equipment outages. This can happen when there are multiple equipment failures, or a single equipment failure at a time when other equipment is out of service for maintenance.

Estimating the average time between interruptions at a point of service is a complex exercise. We used a simplified model to calculate RCP2 targets for the number of interruptions at points of service.11 Sites with N-1 or greater security should expect interruptions only once every few years or less. Sites with N security might expect multiple interruptions per year. This is because the assets that connect customers at N security points of service have single points of failure that can result in an interruption. Reflecting this, we developed individual GP1 targets for each N security point of service.12 Table 2 sets out the RCP2 targets for GP1.

Table 2: GP1 — Number of interruptions

GP1. NUMBER OF INTERRUPTIONS

CATEGORY RCP2 TARGET (COUNT) RCP2 TARGET (FREQUENCY) HIGH PRIORITY 2 • 1 in 10 years

IMPORTANT 9 • 2 in 10 years

9 Events attributed to third parties (e.g. distribution company assets causing auto reclose) are not included in the totals.

10 Commerce Commission. Transpower Individual Price-Quality Path Determination 2015, Part 4 11 This is outlined in Appendix A of the 2013 Service Performance Measures report provided as part of our RCP

2 proposal. An updated list is available on our website. 12 Our approach to setting these N security targets and a full list of the N security targets are given in

Appendix A of our 2013 Service Performance Measures report.

SERVICESREPORT2017

12

STANDARD 26 • 3 in 10 years

GENERATOR 11 • 3 in 10 years

N-SECURITY 56 • Varies by site

GP2 and GP3: long-term duration of unplanned interruptions performance targets The duration of interruptions is largely driven by our operational capabilities. These include how quickly our operators can assess an event and determine the appropriate response, the time our service providers take to get to site and carry out any necessary repairs, and the location of our spares. Tables 3 and 4 set out our GP2 and GP3 targets.

Table 3: GP2 — Average duration of interruptions

GP2: AVERAGE DURATION OF INTERRUPTIONS (MINUTES) CATEGORY RCP2 TARGET (COUNT) HIGH PRIORITY 70

IMPORTANT 100

STANDARD 65

GENERATOR 130

N-SECURITY 80

Table 4: GP3 — Duration of long interruptions

GP3: P90 LONGEST DURATIONS (MINUTES) CATEGORY RCP2 TARGET (COUNT) HIGH PRIORITY 120

IMPORTANT 240

STANDARD 130

GENERATOR 350

N-SECURITY 215

RCP2 Reliability measures performance Table 5 reports on our performance relative to the reliability targets, caps and collars.

Table 5: Grid Reliabi l i ty performance versus Target, Cap and Collar

Number of interruptions (GP1)

Category 2016 2017 Target Cap Collar

High 0 3 2 0 4

Important 2 6 9 4 14

Standard 14 23 26 21 31

SERVICESREPORT2017

13

Generator 4 20 11 6 16

N-security 22 26 56 38 74

GP2: Average duration of interruptions (minutes) (GP2)


High 0 47 70 30 110

Important 14 201 100 30 170

Standard 42 75 65 0 130

Generator 162 105 130 50 210

N-security 167 616 80 45 115

GP3: 90th percenti le duration (minutes) (GP3)


High 0 72 120 80 160

Important 17 482 240 170 310

Standard 116 131 130 60 200

Generator 234 173 350 260 440

N-security 341 1056 215 170 260

Our reliability performance is challenging to predict due to the influence of external factors, such as weather conditions and other events such as fires. Notable events from 2017 include:

• three interruptions to high priority sites, including an interruption at Bromley due to wildfires in the Port Hills near Christchurch.

• 20 interruptions at generator sites. Four of these events were due to heavy snowstorms in the Hawkes Bay, and six were due to birds persistently nesting in equipment serving the Maraetai hydro stations on the Waikato River. We are implementing mitigations for both risks.

• Average duration and 90th duration for important points of supply was affected by the shooting of insulators on the Kaikohe-Maungatapere line that affected supply to Kaikohe

• Average duration and 90th duration for N-security points of supply were due to interruptions at Blackpoint, Culverden and Castlehill. The Blackpoint interruption has heavily influenced the overall results. It was caused by a transformer fault and it took almost 10 days to restore supply. Restoration and performance of N-security sites has been a focus, with a programme of work underway to ensure each site has a plan in place for restoration and continuity.

SERVICESREPORT2017

14

GRIDAVAILABILITYMEASURESThese measures are concerned with the impact that asset availability can have on access to the lowest cost mix of generation.

Both planned and unplanned asset outages can lead to transmission constraints13 and increased system losses.14 However, the impacts vary depending on the circuits concerned. Outages on some circuits can have little or no impact, while outages on others can have significant impacts.

The HVDC circuits allow power flows between the North and South Islands. Power flows from south to north at times when there is an excess of lower-priced generation in the South Island. During times of low hydrological conditions (e.g. dry years) in the South Island, power can be produced by thermal generators in the North Island, allowing the water stored in the South Island lakes to be conserved. HVDC unavailability or constraint may prevent us from achieving optimal dispatch of generation.

Some of the HVAC circuits also have the potential to significantly impact constraints and system losses. We have included an availability measure for selected grid circuits that have significant impacts on either constraints or losses, for example, the 220 kV circuits from Haywards (near Wellington) to Bunnythorpe (near Palmerston North) through to Whakamaru (north of Taupo).

RCP2 Availability measures The two RCP2 grid availability measures are focussed on the transmission circuits that have the greatest impact.

AP 1: HVDC energy availability The HVDC availability measure is the average energy availability of HVDC Poles 2 and 3. The measure is based on the maximum amount of energy the HVDC link can transfer at full loading for a full year, less the reduction in energy transfer caused by de-rating a pole for equipment outages.15

AP2: HVAC availability (selected circuits) The HVAC availability measure is the percentage of time that a selected set of 220 kV circuits is available during a year. We selected circuits based on the potential effect that outages would have on security constraints or system losses (or both).16 The selected circuits are listed in Table 4.

Table 6: Selected circuits for AP2 - HVAC availabi l i ty measure in RCP2

SOUTH ISLAND NORTH ISLAND

13 A transmission constraint is a local limitation in the transmission capacity of the grid. 14 System losses are the difference between the energy injected into the grid by generators and the energy

delivered to final customers. 15 The HVDC link does not operate at full transfer often. However, there are times when it is critical to maximise

the transfer capacity of the link (e.g., extreme hydrological situations in the South Island). The future evolution of the HVDC measure may focus on the actual market effects of de-rating link capacity due to particular equipment outages occurring at a certain time.

16 Our approach to selecting these circuits is covered in more detail in Appendix B of the 2013 Service Performance Measures report.

SERVICESREPORT2017

15

Clyde–Cromwell–Twizel1 and 2 Ohakuri–Wairakei1

Manapouri–North Makarewa1,2 and 3 Te Mihi–Whakamaru1

North Makarewa–Tiwai1 and 2 Bunnythorpe–Tokaanu1 and 2

Clyde–Roxburgh1 and 2 Rangipo–Tangiwai1

Ashburton–Timaru–Twizel1 and 2 Atiamuri–Whakamaru1

Invercargill–Manapouri2 Te Mihi–Wairakei1

Tekapo B–Twizel1 Pakuranga–Whakamaru1 and 2

Islington–Tekapo B1

Ohau B–Twizel3

Ohau C–Twizel4

Ashburton–Islington1

Islington–Livingstone1

RCP2 Availability Targets Our approach to setting targets for both AP1 and AP2 is to assume 100 per cent availability, with adjustments for approved construction outages, efficient preventive maintenance outages and unplanned outages as per GP1 and GP2.17 Our AP1 and AP2 targets for RCP2 are listed in Table 5.

Table 7: AP1 and AP2 — Grid availabi l i ty targets

CATEGORY RCP2 TARGET

HVDC (ENERGY) 98.5

KEY HVAC CIRCUITS 99.6

RCP2 Availability measures performance Table 8 reports on our performance relative to the availability targets, caps and collars.

Table 8: Grid Availabi l i ty performance versus Target, Cap and Collar

Availabi l i ty


HVDC (AP1) 98.9 98.6 98.5 99.5 97.5

HVAC (AP2) 99.0 99.0 99.6 100.0 99.2

17 The approach to setting the availability targets is described in Appendix A of our 2013 Service Performance Measures report.

SERVICESREPORT2017

16

While we achieved our HVDC availability target, performance has been affected by our recent stance to not undertake live line work to ensure we comply with new guidelines that have been published by the Electricity Engineers’ Association. This stance may have some effect on our ability in future to achieve our availability targets.

The HVAC availability target for RCP2 has proven difficult to meet and may be further impacted by our decision regarding live line work. In response, we have worked to an internal target of 98.7% that better balances need for HVAC availability with the need for lines to be de-energised for planned maintenance. We will refine this target for RCP3 so that it better reflects this balance.

ASSETHEALTHMEASURES

Our asset health grid output measures for RCP2 are volumetric delivery targets. Asset health data and models were used as an input into sizing our funding proposal for RCP2, and are used for ongoing prioritisation and scoping. However, asset health is only one input into these processes. Our planners also consider other factors, such as criticality and delivery efficiency, when sizing and scoping our work programmes.

RCP2 Asset Health measures Our Asset health grid output measures show the number of assets in each asset class that, at the time of our 2013 proposal, we expected to replace or refurbish during RCP2. The asset portfolios included in the output measures cover 68% (by value) of our Transmission Lines and AC Station R&R portfolios.

RCP2 Asset Health targets The targets for these measures quantify the number of assets that, in 2013, we expect we would replace or refurbish (i.e. volumetric targets) during RCP2. We have 3 measures with annual targets and 3 measures with targets set over the whole control period.

Table 9: Selected assets with annual Asset Health Grid output volume targets

GRID OUTPUT MEASURE

RCP 2 TARGET 2015/2016 2016/2017 2017/2018 2018/2019 2019/2020

NUMBER OF TOWERS PAINTED (AH1)

451 529 531 553 564

NUMBER OF GRILLAGES COMMISSIONED (AH2)

408 408 408 409 409

NUMBER OF INSULATORS COMMISSIONED (AH3)

1526 1466 1402 1315 1380

Table 10: Selected assets with Asset Health Grid output volume targets for the end of RCP2

GRID OUTPUT TARGET RCP 2 TARGET

SERVICESREPORT2017

17

NUMBER OF OUTDOOR CIRCUIT BREAKERS COMMISSIONED (AH4)

155

NUMBER OF TRANSFORMERS COMMISSIONED (AH5) 25

NUMBER OF OUTDOOR TO INDOOR CONVERSIONS COMMISSIONED (AH6)

16

RCP2 Asset Health measures performance Table 11 reports on our performance relative to the yearly targets, cap and collar set for the measures of:

• the number of towers painted

• the number of grillages commissioned

• the number of insulators commissioned

Table 11: Count of Towers Painted, Gri l lages Commissioned, Insulators Commissioned by Year versus Target, Cap and Collar

Category Year Performance Target Cap Collar

Number of towers painted (AH1)

2016 actual 461 427 467 387

2017 actual 517 523 563 483

2018

forecast 451 517 557 477

2019

forecast 548 558 598 518

2020

forecast 563 555 595 515

Number of gri l lages commissioned (AH2)

2016 actual 271 339 370 308

2017 actual 416 396 427 365

2018

forecast 384 408 439 377

2019

forecast 287 390 421 359

2020

forecast 267 377 408 346

Number of insulators commissioned (AH3)

2016 actual 852 1532 1647 1417

2017 actual 951 1466 1581 1351

2018

forecast 984 1402 1517 1287

2019

forecast 1260 1315 1430 1200

SERVICESREPORT2017

18

Category Year Performance Target Cap Collar

2020

forecast 1360 1375 1490 1260

The grillage encasement strategy has been through a significant review and update since the 2016 AMP and the result of this has been a reduction in the volume of work forecast to be completed in RCP2. The structures and insulators replacement program is constantly reviewed taking into account new condition assessment data and refinements to health modelling. Because of this the forecast for insulator replacement in RCP2 has reduced.

Table 12 reports on our performance against measures AH 4, AH5 and AH6. These measures relate to the whole of RCP2. The number of assets in 2016 and 2017 is reported, along with our forecast for RCP2.

Table 12: Count of Outdoor Circuit Breakers Commissioned, Transformers Commissioned, Outdoor to Indoor Conversions Commissioned for RCP2 versus Target, Cap and Collar

Category 2016-17 RCP2

Forecast Target Cap Collar

Number of outdoor circuit breakers commissioned (AH4)

154 141 153 129 154

Number of transformers commissioned (AH5)

18 26 28 24 18

Outdoor to indoor conversions commissioned (AH6)

14 15 16 14 14

The reduction in the number of transformers commissioned also reflects the use of new planning frameworks and tools that have seen us defer some of this work to deliver cost savings.

PERFORMANCEMEASUREDEVELOPMENTINITIATIVESConsultation for the RCP2 measures identified some additional indicators that were potentially important to stakeholders for which we established (internal) trial reporting during RCP2, with a view to possibly include them as grid output measures in later regulatory control periods.

There are:

• PMD1 – Time to provide initial information following an unplanned interruption • PMD2 – Time to provide updated information following an unplanned interruption

(greater than 30 minutes) • PMD3 – Accuracy of notified restoration times following unplanned interruptions • PMD4 – Extent that Transpower meets planned outage restoration times • PMD5 – Extent that Transpower places customers on ‘N’ security • PMD6 – Number of unplanned momentary (of less than one minute) interruptions • PMD7 – Energy not supplied for each point of service for each unplanned interruption • PMD8 – Extent that Transpower meets planned outage start times for critical circuits

and equipment

SERVICESREPORT2017

19

• PMD9 – Extent that Transpower provides its reports to affected parties on unplanned interruptions within 15 working days of the interruption.

We discuss the findings of our trial and proposed treatment of these performance development measures in Part 3 where we describe our customer and stakeholder engagement and our indicative service performance measures for RCP3.

Asset health measures that are not linked to revenue The following measures represent the forecast change to the average remaining life (a proxy for asset health) based on the impact of RCP2 investment in the asset class and the increasing age of the asset population during the control period.

Table 13: AH1RL – Change in average remaining l i fe, annual targets

GRID OUTPUT MEASURE

RCP 2 TARGET 2015/2016 2016/2017 2017/2018 2018/2019 2019/2020

TOWER COATING OF TRANSMISSION TOWER FLEET

-0.696 -0.565 -0.678 -0.712 -0.697

Table 14: AH4RL - AH5RL – Change in average remaining l i fe, end of control period targets

As noted earlier, we obtained an exemption from reporting on these measures in 2016 so that we can develop improved measures that align with our evolving asset health models and provide more meaningful information for stakeholders.

ASSET HEALTH GRID OUTPUT MEASURE RCP 2 TARGET

OUTDOOR CIRCUIT BREAKER FLEET -0.258

TRANSFORMER FLEET -0.194

SERVICESREPORT2017

20

P A R T 3 : S T A K E H O L D E R E N G A G E M E N T A N D T H E R E F R E S H O F M E A S U R E S F O R R C P 3

In this Part we:

• summarise the customer and stakeholder engagement we carried out to support refreshing our service performance measures for RCP3

• set out our proposed measures for RCP3 • the development of our asset health measures for RCP3.

We are currently in the process of refreshing and evolving our service performance measures for RCP3. Our approach has involved seeking customer and stakeholder feedback throughout the formulation of proposals. We are appreciative of the feedback so far and for the support we have received.

ENGAGEMENTPROCESSOur customer and stakeholder engagement process is outlined below. We have been very conscious of the extensive involvement many of our customers and stakeholders have in current regulatory processes and consultations. We have received supportive feedback on the process including that it has struck the right balance. Each step in the process below builds on the preceding work.

To date, we have:

1. Published the November 2016 paper: The 2016 paper sought feedback on the engagement process and areas where further evolution of the service performance measures might be warranted. It also provided additional supporting material for those new to this subject area.

2. Held the February 2017 focus group meeting: A focus group of representatives from a cross section of our customers, stakeholders and end consumers. The focus group discussed possible evolution of the measures and provided its views on what works in practice and what doesn’t.

3. Published the Apri l 2017 paper: This paper provides more detail on the proposed RCP3 service performance measures, and the proposed methodologies by which performance against the measures will be calculated. It also sought formal written feedback.

4. Presented at Apri l 2017 industry workshops: We presented the key proposals at industry workshops in Auckland, Wellington and Christchurch. These workshops included representatives of our customers and other stakeholders.

5. Presented at Apri l 2017 regional seminars: We presented the key proposals at regional sessions hosted by our direct customers, Unison (Hastings) and Delta (Dunedin).

SERVICESREPORT2017

21

Feedback The measures proposed for RCP3 (and presented here) have evolved based on feedback received through the engagement process. The improvements have been iterative, based on feedback at each stage. The submissions received along with a summary of submissions at each stage can be found here.18

RELIABLEGRIDMEASURESFORRCP3For RCP3 we propose to retain the rel iable grid measures GP1 and GP2, which establish the reliability experienced by five defined point of service categories. We propose to discontinue GP3, the P90 reliability measure. In practice, the number of events in several of the categories is too small for a robust P90 calculation. Moreover, as large events are very rare we do not consider there is sufficient baseline data for the P90 value to be meaningful. Experience has also shown that during very long outages specific circumstances typically drive the outage length.

Review of point of service categorisation In RCP2 points of service were allocated to five categories that reflected the relative economic impact of supply interruption.

Feedback suggested that for RCP3 we should review both the categories and the methodology for assigning points of service to categories.

We are currently undertaking this review and will consult with customers and stakeholders on the revised methodology, categories and final point of service allocations prior to our RCP3 submission. Part of this review includes a review of value of lost load (VoLL) which is a key input into the categorisation review.

GRIDAVAILABILITYMEASURESFORRCP3For RCP3 we propose to retain both grid avai labi l i ty measures, AP1 HVDC availability and AP2 HVAC availability.

The purpose of the grid availability measures is to incentivise us to ensure the Grid is available as much as possible, minimising losses and constraints that will impact the electricity market. Within this context the key issue identified within our engagement process to date is that AP2 does not incentivise efficiency and optimisation in scheduling planned outages on network equipment. We believe that this feedback is valid.

In considering alternatives to AP2 that may better achieve its intent, it is evident that many mechanisms are susceptible to significant influence from factors outside our control. Accordingly, we are currently undertaking an outage planning review to further improve our processes including how we communicate and the significant work we do to determine the trade-offs, timing and efficiency that occurs within the planning horizon.

18 https://www.transpower.co.nz/transpower-service-level-refresh-rcp3

SERVICESREPORT2017

22

Review of HVAC circuits Feedback has indicated a need to review the 27 220 kV circuits monitored for the AP2 measure and whether additional circuits should be added to give a better ‘220 kV Grid backbone’ representation. We will consult on the findings of this circuit review at our next consultation.

Inclusion of ancillary equipment in AP1 measure We received feedback from some generators that the effectiveness of the AP1 HVDC availability measure is restricted by the exclusion of filter bank equipment. Filter bank equipment can that impacts on the reserve capability of the HVDC. While we have considered changes to the HVDC measure methodology to account for this, we are currently working on a HVDC control system change with our provider to remove this restriction. We expect to implement this change during an outage later this year. As such we do not feel modifications to the methodology for RCP3 are going to be necessary.

One generator also raised that other ancillary equipment (in addition to filter banks) limits the HVDC power limit which is correct. However, much of this is dependent on the Wellington demand. For example, under some Wellington demand scenarios a Haywards STATCOM outage will have no effect on the HVDC power limit, but under other Wellington demand scenarios this can reduce the HVDC limit by 200 MW. As Wellington demand is outside our control and the HVDC power limits depend on a complex interaction of factors, we do not feel it reasonable to include ancillary equipment in the measure.

A new ‘Adherence to the outage plan’ measure Both feedback from stakeholders and our own experience indicates that market impacts are less likely where outages are undertaken within their scheduled timeframes. Adhering to our outage plans is obviously important: it allows market participants to have confidence to arrange and execute appropriate risk management mechanisms and for traders to set up book positions in advance, leading to better long term outcomes for end-consumers (relative to uncertainty and volatility being induced by transmission outages being moved at short notice or shifting and running over the scheduled end time). It also means customers can plan to minimise inconvenience and disruption. Based on this we propose introducing a third grid availability measure around adherence to the outage plan. Feedback we have received is that this should be biased against late restoration only as early restoration almost always has positive impact for consumers.

We are currently investigating this feedback, how we might implement such a measure and any changes to our tools or process that may be required to do so. We will engage with stakeholders on our findings as part of our next consultation.

EVENTCOMMUNICATIONSMEASUREFORRCP3For RCP3 we do not propose setting any formal review-linked event communication measures. Rather, we propose to continue with our post-event surveys of customers. Since October 2015 we have been undertaking post-event surveys of our customers. Although data from these indicate that customers are satisfied with our service and communications, the value lies in the comments. These comments show trends as well as specific areas of improvements. The comments provide a platform for us to have open and honest conversations with our customers about the events and projects.

SERVICESREPORT2017

23

For these reasons, we have received strong support to continue this process in lieu of developing any of PMD1, 2 or 3 into formal revenue-linked measures We believe that a survey approach serves the dual purpose of providing more meaningful feedback on the quality of communication, and how communication may be improved in future. This approach also allows a fuller range of communication factors to be considered, such as mode of communication and information clarity. Communication will continue to focus on our customers, as opposed to wider end customers.

TREATMENTOFPERFORMANCEDEVELOPMENTINITIATIVESFORRCP3We recognise the value both to customers and to ourselves in better understanding different aspects of our performance. When these development measures were set we were not fully collating this information. As such, these measures were set as trial development measures to understand whether the indicators could be meaningful and potentially adopted for use as formal measures in RCP3. We internally trialled these development measures for one year, and engaged with customers and stakeholders on whether to adopt them as future measures for RCP3. A summary of each of the nine indicators is provided below.

PMD1 – Time to provide initial information following an interruption We collated data on the time from the unplanned interruption to the initial response contact to a direct customers’ duty person. We set a target of 15 minutes and internally reported the percentage of times this target was met. For the trial, all contacts were made within the target time, with a maximum time of 8 minutes.

While timeliness of communication is important and this indicator has the advantages of being readily measurable and objective, feedback received was that this indicator was not useful nor did it necessarily gauge the usefulness of the information to the customer. Similar positives and negatives were identified with the other event communication measures trialled in RCP2 and which are described further below. There was support from our engagement that current communications approaches are working very well.

We have received support to not adopt this measure for RCP3.

PMD2 – Time to provide an update (for extended interruptions) Most of our interruptions are restored within 30 minutes, but the intent of this measure was to incentivise us to communicate with customers who experience longer interruptions. The target for providing an update was within 30 minutes of the initial communication. For the period of the trial all updates were provided within the target time.

As with PMD1, this indicator is readily measurable and objective, but it does not necessarily gauge the usefulness of the information to the customer.

As with PMD1, we received support to not adopt this measure for RCP3.

PMD3 - Accuracy of restoration information For the trial we recorded the percentage of unplanned interruptions that were restored:

a) within 10 minutes of the advised estimated restoration time b) within 30 minutes of the advised estimated restoration time

SERVICESREPORT2017

24

c) more than 30 minutes after the advised estimated restoration time.

The data indicated significant variability in the accuracy of forecasts about restoration times. This is not unexpected because it can be challenging to identify event causes prior to the initial communication. Likewise, it can be difficult to predict the remediating steps required, and the time needed to action them. There will always be room for improvement in this area.

Accuracy of predicted restoration time is clearly important to customers. In this respect, PMD3 is arguably one of the more valuable potential development measures. However, if it was used for performance evaluation purposes, it could create perverse incentives. For example, it could encourage us to provide unduly long estimated restoration times or to slightly delay restoration in some cases so that a revised (and very ‘accurate’) customer communication update could be issued. Feedback was very clear that the quality and usefulness of the information received was the most important.

We have received support to not adopt this as a measure for RCP 3.

PMD4/ PMD8 – Extent that Transpower meets planned outage start and end times Historical experience indicates that market impacts (and associated economic effects) are less likely where outages are undertaken within the scheduled timeframes. For this reason, the proposed indicator appears useful to generators and customers. It also seems reasonable to expect that we will have better (albeit not perfect) knowledge about likely restoration times for planned outages than for unplanned outages.

This measure has the potential to create perverse incentives because it encourages conservative planning which extend actual outage times, hedging and, ultimately, costs. It also could incentivise us to proceed with a planned outage at a scheduled time, even if there are factors which mean it should be delayed. Likewise, it could deter us from undertaking work outside a plan even where that is unlikely to have a material market impact and is better for all parties involved. Caution was also raised about the safety of staff taking unnecessary safety risks to return assets to service within planned timeframes.

Feedback supports a measure around adherence to the outage plan. We propose investigating the adoption of a modif ied version of PMD4 and 8 as a potential new grid avai labi l i ty (AP3) measure for RCP3.

PMD5 – Extent to which customers are placed on N security Reporting time on N security would allow the assessment of how often customers are on N security and how time on N security influences risk of non-supply. The latter could also inform work on how to reduce the risk of non-supply.19

19 Transpower, submission to Commerce Commission, 2013.

SERVICESREPORT2017

25

Our trial reporting on this has proven to be relatively resource intensive (requiring one person to dedicate one week each month to the preparation of statistics). We believe the rationale for this information remains valid, but the usefulness of this to customers and our own decision-making needs to be further explored. Feedback from customers and stakeholders supported the view that PMD5 was not appropriate as a future performance measure and there is insufficient information to identify an optimal or target level, let alone derive incentive rates. In addition, the GP1 grid reliability measure should provide insights over time as to the underlying trend in levels of security (i.e. all things being equal, the more often a point of service is on N security, the more frequent loss of supply events will be).

We have received feedback that suggests not adopting this measure for RCP3, We do propose to continue conversations with customers around how we can provide this type of information in a more streamlined way to those who are interested.

PMD6 - Number of unplanned momentary interruptions Under RCP2, unplanned momentary (less than one-minute duration) interruptions are excluded from the GP1, 2 and 3 measures. This is because many of these interruptions are caused by factors outside our control, such as lightning strikes. Thus, a particularly stormy period or year, or conversely a very calm year, could skew results. In addition, for many customers, a momentary interruption (e.g. for an auto reclose to operate) will not impose material costs.

Notwithstanding their exclusion from the primary reliability statistics, RCP2 included such interruptions as a development measure. Our feedback has questioned the value of adopting this measure for RCP3. There was a common view that service performance measures should reflect situations that are predominantly within our control.

The key feedback through our engagement was the sharing/ reporting of this information. This ensures the industry is aware that momentary interruptions are being considered. During RCP2 we established a voltage disturbance working group in response to requests from customers to better understand voltage disturbances on our network following discussions about the proposed momentary interruptions measure. The initial focus of the working group is around the development of methodologies and resources to measure, characterise, record and report voltage disturbances at selected grid exit points.20 Customers support not continuing with this measure because the topic will be discussed on an on-going basis at the working group with the expectation that information around these interruptions will be openly reported.

We have received support to not adopt this as a measure for RCP3.

PMD7 - Energy not supplied for each point of service for each unplanned interruption

We trial-reported data on energy not supplied per interruption event at the point of service level for a period of one year during RCP2.

20 Those interested in knowing more or participating in this working group should contact Stephen Jones, Strategic Asset Manager. [email protected]

SERVICESREPORT2017

26

Point of service level interruption data is already made available via Information Disclosure provisions, but not at the event level.21 There was general agreement from our consultation feedback that this was not appropriate as a service measure. More timely information that is updated regularly would be more valuable. We are interested in working with customers to establish precisely what customers and consumers would find value in us reporting.

We have received support to not adopt this as a measure for RCP3.

PMD9 - Extent that Transpower provides its reports to affected parties This indicator is about us making post-event reports to affected parties within 15 working days of an interruption. This recognises that customers will have a clear interest in the underlying causes of interruptions, because this may help them to assess the future likelihood of similar events, and inform any actions they take because of an event.

However, it is not clear that the proposed indicator is necessary. We already have obligations, under connection contracts with our customers, to provide post-event reporting.22 These provide a 42-day window, a length of time we consider more appropriate given the requirement to gather, analyse, and write-up technical information. It is not clear why shortening this to 15 working days would yield any material customer benefit, given that the immediate outage will have ended some time earlier. Participants at our forum did request that we aim to improve our post-event reporting with clear outcomes and actions.

Feedback raised no objections to keeping this report ing at the exist ing 42-day period so we wil l not be adopting this measure for RCP3.

SUMMARYOFCHANGESTOSERVICEMEASURESFORRCP3

A summary of current RCP2 measures and proposed changes for RCP3 is provided in Figure 4.

21 See schedule G4 of Transpower Information Disclosures. 22 The Benchmark Agreement, which is a default connection contract set out in the Electricity Industry

Participation Code, includes an obligation for us to provide customers with a post-event report within 42 days of an unplanned interruption.

SERVICESREPORT2017

27

Figure 4: Changes to service performance measures from RCP2 to RCP3

Category Performancemeasure RCP2code

ProposeasameasureforRCP3?

Reliability Numberofunplannedinterruptionseachyearbycustomercategory

GP1 ü

Averagedurationofunplannedinterruptionsbycustomercategory(min/yr)

GP2 ü

Durationof90thpercentileunplannedinterruptionbycustomercategory(min/yr)

GP3 û

Availability %availabilityofHVDC AP1 ü

%availabilityofselectedHVACcircuits AP2 üCategory Developmentmeasure RCP2

codeProposeasameasureforRCP3?

CustomerService/EventCommunications

Timetoprovideinitialinformationfollowinganunplannedinterruption PMD1 ü


Timetoprovideupdatedinformationfollowinganunplannedinterruption(greaterthan30minutes)

PMD2 û


Accuracyofnotifiedrestorationtimesfollowingunplannedinterruptions PMD3 ü

Availability ExtentthatTranspowermeetsplannedoutagerestorationtimes

PMD4 üAvailability ExtentthatTranspowerplacescustomerson‘N’

securityPMD5 û

Reliability Numberofunplannedmomentary(oflessthanoneminute)interruptions

PMD6 ûReliability Energynotsuppliedforeachpointofservicefor

eachunplannedinterruptionPMD7

Availability ExtentthatTranspowermeetsplannedoutagestarttimesforcriticalcircuitsandequipment PMD8 ü

Reliability ExtentthatTranspowerprovidesitsreportstoaffectedpartiesonunplannedinterruptionswithin15workingdaysoftheinterruption.Transpowerwillreportanyexpectationsonthenumberoftimesitdidnotmeetthetimeframe

PMD9 û

CustomerService/EventCommunicationsExistingposteventsurvey.Focusesontimelyinformationprovisionandcommunications.

CS1

IncorporatesPMD1,2and3

includingqualitative

informationonthequalityofthe

communications

Otherreporting:finalisationofposteventreports,onlineupdatesforenergynotsuppliedforeachPOS,voltagedisturbanceworkinggroup

CurrentRCP2Measures ProposedRCP3Measures

Category Developmentmeasure RCP2code






PMD2 û





securityPMD5 û






PMD9 û

Category Developmentmeasure RCP2code






PMD2 û





securityPMD5 û






PMD9 û



Reliability Numberofunplannedinterruptionseachyearbycustomercategory

GP1 ü

Averagedurationofunplannedinterruptionsbycustomercategory(min/yr)

GP2 ü

Durationof90thpercentileunplannedinterruptionbycustomercategory(min/yr)

GP3 û

Availability %availabilityofHVDC AP1 ü

%availabilityofselectedHVACcircuits AP2 ü


Proposedrefinement

Numberofunplannedinterruptions eachyearbycustomercategory

GP1 Categorisationtobealignedwith

VoLL

Averagedurationofunplannedinterruptions bycustomercategory(min/yr)

GP2 Categorisationtobealignedwith

VoLL

%avai labi l i tyofHVDC AP1 -

%avai labi l i tyofselectedHVACci rcuits AP2 Circuitsincludedtobereviewed

AdherencetotheoutageplanExtentthatTranspowerkeeps toplannedoutagetimes

AP3? Measureandtargettobedeveloped

Rel iabi l i ty

Avai labi l i ty

SERVICESREPORT2017

28

DEVELOPMENTOFASSETHEALTHMEASURESFORRCP3

We propose to replace the current Asset Health Grid Output measures with alternative asset health measures for use in RCP3. We are developing measures that will more accurately reflect how we utilise asset health information in our business. To achieve this, we have been developing an asset health report to pilot for the remainder of RCP223. The measures used in this report will form the basis of our proposed RCP3 asset health revenue linked output measures.

The process we have taken to develop an alternative set of asset health measures

During RCP2, we undertook an improvement initiative to develop our asset health models, data and processes24 that use AHI as inputs in our asset planning decisions. We first developed asset health models in RCP1, and used them to help develop our refurbishment and replacement plans for the current IPP. Our understanding of how to model and use asset health continues to grow and develop.

We have made several improvements to the way we use asset health including: improvements in modelling such as automating models and developing more mature models for certain asset classes; improving the way we collect and assess asset condition data; as well as increasing the coverage of assets that we have health models for.

We intend to publish a report on asset health in October 2017 that will incorporate these latest developments. The report will show the population of selected assets with asset health scores indicating the current health of the asset class. We also intend to illustrate how our replacement and refurbishment programme will affect the health profile of our assets over the remainder of the control period.

We have also been working with the Commerce Commission to develop an improved asset health reporting methodology that could be used to develop revenue linked output measures for RCP3. These measures will incentivise us to deliver on what matters, the health or condition of the assets. This contrasts with the current asset health measures which incentivises the delivery of a fixed asset replacement plan. While the current regime provides a more objective and certain target, it does not support the flexibility needed to provide the best outcomes. We have been engaging with the Commission on whether the current grid output framework is a suitable mechanism for our proposed asset health output measures. Our preferred approach to linking asset health output measures to revenue incentives is a non-mechanistic approach where we would agree up-front a set of asset health outcomes and the circumstances under which incentives would or would not apply. For example, the incentive regime should not supress bad news (e.g. about asset condition, deterioration rates or failure modes) or encourage inefficient interventions (e.g. replacing low-risk assets) but should penalise simple failure to deliver. Next steps

23 As discussed in Part 1, this asset health reporting pilot would replace our obligation in the current IPP to report on non-revenue linked asset health measures based on average remaining life.

24 March 2016 Initiatives Plan Update: https://www.transpower.co.nz/sites/default/files/plain-page/attachments/Regulatory%20Initiatives%20Plan%20-%20March%202016%20Update.pdf

SERVICESREPORT2017

29

Our first major milestone is to agree a set of alternative asset health pilot reports with the Commerce Commission to replace the current reporting obligations for RCP2. We will consult with interested stakeholders in the second part of 2017 on the new asset health report as well as on our views of potential RCP3 measures and the framework for linking these measures to revenue. As mentioned above, we do not believe a single point target and cap and collar regime would be suitable for asset health measures. There will need to be amendments to the Capex Input Methodology if we were to implement our proposed framework.

We are engaging with the Commerce Commission in their current review of the Capex IM.

SERVICESREPORT2017

30

P A R T 4 : H O W W E U S E T H E T A R G E T S A N D F U T U R E D E V E L O P M E N T

This Part:

• summarises the role of service measures and targets in our business • describes our current practices and development plans.

THEROLEOFSERVICEMEASURESANDTARGETS

At a high level, we use service measures and targets as follows:

• To discuss our services with customers. The measures should reflect what is important to customers. Having clear measures helps support discussions with our customers, including commercial and operational negotiations.

• As a critical input into to asset management decisions. Asset management interventions should be directed towards meeting our service targets.25 Having clear measures helps align and prioritise our efforts.

• To shape our expenditure plans. Having clear service targets helps us to drive towards the least-cost means of meeting service expectations.

Current practices As a service-focused organisation we use our services framework (which includes the three services most relevant to price-quality path regulation) to help frame our conversations with customers.

Combined with key principles and strategic goals, our services framework is reflected through our strategies and applied within our planning processes. Planning occurs within our business on a continual basis, with ongoing assessment of investment needs, capability requirements, and deployment and delivery planning.

We use point of service categories to help prioritise our work. We use service categories to map criticality scores against individual assets. For some asset types the criticality score uses services categories in conjunction with a safety score. This is helping to more systematically incorporate criticality into our risk assessment and prioritisation processes.

The service measures are an important part of our operations. We use the targets to inform our contracts with service providers and are using the reliability and availability targets in our works planning processes.

25 In practice, interventions are also directed towards other objectives, such as worker and public safety, environmental compliance or lifecycle cost optimisation.

SERVICESREPORT2017

31

Since 2014 we have been using the service performance measures for internal performance reporting. This process is helping to embed the measures within our business. Since July 2015 the service targets have been embedded in our Statement of Corporate Intent and internal measures of success. We report on the measures in our quarterly NZX operational performance disclosures.

Our performance against the current service performance measures can be found in Appendix B.

Future development – context As new technologies and business models emerge, electricity consumers will have more choice about when and how much wholesale electricity to consume. Consumers will demand different services from us than we provide now. Because some of these technologies will come from the mass market we can also expect that these differing services may be fast-changing themselves.

These changes will bring a need for fundamental changes to the existing electricity industry. In 2016 we developed Transmission Tomorrow to provide a perspective on what these new technologies and business models might mean for our business and the wider electricity industry. At a fundamental level, we foresee that our own business may change over time from providing 24/7 reliability of supply and real-time balancing of supply and demand, to providing a resilient battery-charging service. Our priority could change from achieving very low failure rates in our network to ensuring we can restore supply in a timely manner.

However, a key take-out from our Transmission Tomorrow work is that while the power system of tomorrow will look very different from the power system of today, the grid will still have a vital role to play. We will still be in the business of safely and carefully managing long-lived assets and complex systems that provide essential services to New Zealand communities, households and businesses and so will continue to focus on reducing costs and becoming more innovative while sustaining a focus on good asset stewardship and risk management.

Key to our future success will be to ensure we:

• understand future electricity consumer requirements and preferences

• understand the pros, cons and costs of the choices consumers will have

• have a good understanding of the relationship between our costs and services.

Future development – framework As we further develop our asset management capability, we are exploring to what extent we can integrate services into the various “layers” of our network development, maintenance and operations. The following diagram illustrates how these layers stack together to deliver the reliability outcomes experienced by customers.

SERVICESREPORT2017

32

To date our main focus area has been developing a criticality framework that can bring service objectives into our replacement planning (alongside other drivers). We have also been exploring how service objectives could integrate into configuration decisions and we have been working with customers on mitigation plans and response capability for N-security sites.

We will work on maturing the way service objectives integrate into and across the other layers alongside maturing our service measures and targets.

Configuration • Capacity • Redundancy (N,

N-1, etc)

Built Standard • Strength • Durability • Quality

Replacement • Prioritisation

Maintenance •  Inspection • Servicing • Repair

Contingency • Spares • Mitigation plans

Operations • Operating

redundancy • Diagnostics • Response

capability

SERVICESREPORT2017

33

A P P E N D I X A : T Y P E S O F G R I D O U T P U T M E A S U R E S I N T H E C A P E X I M

The Capex IM refers to a number of different types of ‘Grid Output Measures’. The box below summarises these terms and explains how our service performance measures align with these terms.

TYPES OF ‘GRID OUTPUT MEASURES’ (CAPEX IM)

TYPE OF MEASURE

CAPEX IM DEFINITION OUR RCP2 MEASURES

GRID PERFORMANCE

‘measure that quantifies the level of service received by consumers’.

• Our Reliable Grid Measures GP1, GP2, GP3.

ASSET PERFORMANCE

‘measure that quantifies the performance, reliability or availability of the grid...’ This can be at the level of individual assets, an aggregation of assets or the grid.

• Our Grid Availability Measures AP1, AP2.

ASSET CAPABILITY

‘measure that quantifies the capability or utilization of the grid …’ and ‘reflects the output or benefit (where benefit may include a reduction in risk) delivered by investment in additional capacity to supply electricity transmission services.’

• We have not proposed any asset capability measures.

ASSET HEALTH ‘measure that – (a) quantifies the fitness for service of the grid …’ ‘and (b) reflects the output or benefit (where benefit may include a reduction in risk) delivered by expenditure….’

• We have not proposed asset health measures as Grid Output Measures, although we are developing and using asset health indices to prioritise our capital expenditure.

OTHER ‘GRID OUTPUT MEASURES’

Not defined • We have not proposed any other ‘Grid

Output Measures’.

SERVICESREPORT2017

34

A P P E N D I X B : O U R C U R R E N T P E R F O R M A N C E A G A I N S T T H E S E R V I C E P E R F O R M A N C E M E A S U R E S

The combined incentives for GP1 (number of interruptions) were marginally positive at $0.03M. The number of interruptions was less than the target for important, standard and N security service performance categories resulting in positive incentives. However, the high priority and generator category customers have experienced more interruptions than the target resulting in negative incentives.

The combined incentives for GP2 and GP3 (average and P90 restoration durations) were marginally negative at -$0.15M. The major contributing events to N-security GP2 and GP3 measures are Blackpoint, Culverdeen and Castlehill interruptions. The major contributor to the Important GP2 and GP3 measures was the KOE-MPE insulator gunshots event.

The AP1 HVDC measure was above target. The AP2 HVAC measure was below target and this was due to more planned maintenance than originally envisaged.

RCP2PerformanceMeasuresSummary-Jun-17(Final)

Target Cap Collar

GP1:Numberofinterruptions(p.a.)HighPriority 21 3 2 0 4 3 ● 606 -0.61 -0.61Important 39 6 9 4 14 6 ● 242 0.73 0.73Standard 79 23 26 21 31 23 ● 133 0.40 0.40Generator 39 20 11 6 16 20 ● 133 -0.67 -0.67N-security 44 26 56 38 74 26 ● 10 0.18 0.18

GP2:Averagedurationofinterruptions(min)HighPriority 21 47 70 30 110 47 ● 15 0.35 0.35Important 39 201 100 30 170 201 ● 9 -0.63 -0.63Standard 79 75 65 0 130 75 ● 5 -0.05 -0.05Generator 39 105 130 50 210 105 ● 4 0.10 0.10N-security 44 616 80 45 115 616 ● 3 -0.11 -0.11

GP3:P90Longestdurations(min)HighPriority 21 72 120 80 160 72 ● 15 0.60 0.60Important 39 482 240 170 310 482 ● 9 -0.63 -0.63Standard 79 131 130 60 200 131 ● 5 -0.01 -0.01Generator 39 173 350 260 440 173 ● 4 0.36 0.36N-security 44 1056 215 170 260 1056 ● 3 -0.14 -0.14

Availability(%)AP1:HVDC - 98.64 98.5 99.5 97.5 98.64 ● 1000 0.14 0.14AP2:HVAC - 98.98 99.6 100.0 99.2 98.98 ● 2500 -1.00 -1.00

UnAvailability(durationhours)4AP1:HVDC(pole-hours) - 237.9 263 88 438 238AP2:HVAC(cct-hours) - 2405.4 949 0 1897 2405.4

YTD 12mthTotal -$0.97m -$0.97m

1 Accumulativefromstartoffinancialyear2 FromTranspowerIndividualPrice-QualityPathDetermination2015,dated28Nov2014.3 RCP2Targetdividedby12MonthRolling.Amberif12mthwithin10%oftarget;redifgreaterthantarget.4 AP1convertedtototalFTEhoursfortwopoles;AP2convertedtototalhoursfor27circuits

12MonthRollingTrafficLight3

Incentiverate2

($kperunitfromtarget)

YTDValueof

Incentive($m)

IncentiveCalculatedon12mthrolling($m)

Measure Category Numberofsites

2016/17ActualYTD1

RCP2Targets2 12MonthRolling

TRANSPOWER NEW ZEALAND LTD THE NATIONAL GRID

WWW.TRANSPOWER.CO.NZ

services report a supporting document to … · services report a supporting document to our...

Documents