CUSTOMER-FACING GRID PERFORMANCE MEASURES

Size: px
Start display at page:

Download "CUSTOMER-FACING GRID PERFORMANCE MEASURES"

Transcription

1 CUSTOMER-FACING GRID PERFORMANCE MEASURES CONSULTATION PAPER Transpower New Zealand Limited October 2012 CUSTOMER-FACING GRID PERFORMANCE MEASURES Transpower New Zealand Limited All rights reserved. i

2

3 Chapter 1: Background information Table of Contents Summary and request for feedback... 2 Meetings with customers and consumer representatives...2 Request for feedback Background information Introduction Grid configuration and the price service trade-off Where our grid performance measures and targets fit Commerce Commission regulation How we will use our customer-facing performance targets What performance matters to customers? Number and length of interruptions are important Measures and targets should be meaningful to customers; system averages are of limited relevance Customer expectations vary Customers tolerance of events depends on the cause and how good we are at providing information Overall performance matters; grid performance is only part of the overall service delivered to consumers Proposed measures and targets Different target categories one size does not fit all Number of power outages Duration of power outages Information provision during power outages Transmission outages relative to distribution Your feedback and next steps A.1 Feedback form

4 Chapter 1: Background information Summary and request for feedback We are introducing new measures and targets for grid performance based on our view of what performance matters to our customers and end consumers. These targets will be used to prioritise our expenditure and to inform the revenue incentive regime applying to Transpower in our next regulatory control period ( ). We have prepared this proposal and now want to gain input from our customers and consumer representatives. Our existing grid performance measures are average high-level measures such as HVAC circuit availability and total impact of interruptions. The main differences between these existing measures and our proposed customer-focused measures are: - a change in emphasis from measures that are focused on the performance of our assets to measures of the service received by our customers - different targets reflecting different categories of customers at grid connections - forward-looking targets that reflect what customers expect rather than targets based solely on historical performance. Meetings with customers and consumer representatives This paper sets the scene for more detailed discussions. We plan to meet with groups of our customers and stakeholders. These meetings will provide an opportunity to discuss the proposed measures and targets in more detail and to gain feedback from customers. Request for feedback We welcome feedback from our customers and consumer representatives. This paper includes a feedback form (Appendix A.1). So we can consider your feedback we need to receive it by 5 December Please send your feedback to: Graeme Ancell, Power System Planning Manager, Transpower at: PMFeedback@transpower.co.nz. 2

5 Chapter 1: Background information 1 Background information 1.1 Introduction We are putting in place a programme of initiatives to improve the way we manage and invest in our assets. As part of this we are developing a set of customer-facing grid performance measures that will guide our planning, target our expenditure and improve our management of the grid. We already use grid performance measures. The measures are central to the way we run our business, but we want to refine these measures for two reasons. First, we think there are benefits from introducing measures that reflect what matters to both our customers and to end consumers. This is a change in emphasis from measures that are focused on the performance of our assets to measures that reflect the service customers receive. This will help us to target our capital and maintenance expenditure to deliver performance across the grid that our customers value. Second, the Commerce Commission has introduced a regulatory regime with stronger links between our revenue and service (such as reliability) rather than inputs (such as asset replacement). We support this and as part of this initiative are working with the Commission to link our performance, based on the new performance measures, to our revenue to ensure we are investing efficiently. Developing more customer-focused measures and targets is a key part of this. This paper describes the possible customer-facing performance measures and how we will set targets for these measures. We invite customers and stakeholders to provide feedback. We plan to meet customers to discuss these ideas and possible targets in more detail as part of this consultation. A note on terminology Interruptions, outages, loss of supply In this paper we use the term outage to mean an event that causes a loss of supply or loss of generation connection. We are using the term outage because it is more relevant to consumers than interruption and is often used by distributors in communications with their customers. Outages can be unplanned or planned. These performance measures are concerned with unplanned outages only. Customers and consumers In this paper customers refers to our customers the generators, distributors and large consumers directly connected to the grid. Consumers refers to end users. Quality or performance measures and targets We use the term measures to mean a metric or way of measuring performance (such as number of outages) and target to mean the objective for which we should aim (such as less than 10 outages per year). The customer-facing measures and targets we discuss here refer to the service our customers receive at their grid connections. 3

6 Chapter 1: Background information 1.2 Grid configuration and the price service trade-off Ideally, consumers would be able to choose the level service they require from the transmission grid for the price they are willing to pay. Consumers would weigh up the cost and inconvenience of possible interruptions to their power supply against the increased cost of increased service. However the nature of both transmission and distribution networks means the same assets are often used to supply many consumers. Therefore aside from a few large consumers with dedicated grid connections, we cannot provide different levels of service for each individual consumer. The service consumers do receive is to a large extent determined by the transmission assets employed to deliver electricity to their grid connection. The way the grid is designed and which assets are used, already reflects a trade-off between cost and service. The trade-off is made either through customer choice for dedicated connection assets or through regulated investment decision making frameworks on the main interconnected grid (ie the Grid Reliability Standards). We have a grid with built in redundancy so we can maintain electricity supply when assets fail or require maintenance. The level of redundancy ensures that the loss of a single major item such as a transmission circuit or transformer will not result in an outage to customers (this is commonly referred to as N-1 security) 1. Most customer grid connections have an N-1 security. This means they pay for duplicated assets at the grid connection to avoid outages from unexpected equipment failures and allow maintenance to occur without an interruption to supply. Some grid connections have a lower level of security where some major assets are not duplicated (N security). Customers endure more outages as a result. This can be as the customer has identified that the benefit from increasing security by installing more assets does not justify the added cost. The service customers receive at the grid connection is determined not only by the assets employed, also by how well we manage and maintain those assets and how well we deal with outages when they occur. It is our abilities in these two areas which the performance measures discussed in this consultation seek to address. 1.3 Where our grid performance measures and targets fit As a monopoly transmission company the service we provide and the revenue we can recover from our customers is regulated. The Commerce Commission sets our overall revenue allowance and approves our performance or quality targets. 1 In New Zealand, we generally do not consider triplicated assets (N-2 security) to be economic and customers tolerate some outages from multiple equipment failures as long as they occur infrequently. 4

7 Chapter 1: Background information Transpower s revenue is recovered from our customers through the Transmission Pricing Methodology 2. There is no direct link between the performance a customer receives at the grid connection and the price paid. Within this regulated environment our grid performance measures have dual roles: providing information to enable regulators and customers to assess whether our performance is appropriate; and, encouraging us to focus our spending on our assets efficiently, where the benefits to customers justify the costs. Our existing high-level grid performance measures are based on overall system wide events and averages, they are: HVAC circuit unavailability (unplanned) Total impact of interruptions (in system minutes) Loss of supply event frequency number of loss of supply events > 0.05 system minutes Loss of supply event frequency number of loss of supply events > 1 system minute HVDC bi-pole unavailability (unplanned). Most measurements of international transmission system performance are based on these types of measures. They do not reflect the differing performance individual customers receive or expect. In addition, the targets set for these measures generally reflect improvements on historic performance rather than what the assets are capable of, or the desired long term performance required at each grid connection. 1.4 Commerce Commission regulation The form of price-quality regulation that the Commerce Commission has applied to Transpower since 1 April 2010 is called individual price-quality path regulation. The Commission made two determinations that set Transpower s price-quality path in December 2010 and January These cover a four year period to March 2015, a period known as RCP1. They set Transpower s revenue, provided some incentives and set performance measures and targets (the Commission refers to these as quality standards). Our existing performance targets for RCP1 are derived from the historic performance of the grid. For our next regulatory control period (RCP2, April 2015 to March 2020), the Commission will set a new price-quality path that will link a portion of our regulated revenue to our performance. In advance of this period we must provide a proposal detailing our proposed expenditure, performance measures and targets and an incentive regime 3. 2 The Transmission Pricing Methodology, schedule 12.4 of the Electricity Industry Participation Code The Electricity Authority is reviewing the methodology. 3 Commerce Commission, Transpower Capital Expenditure Input Methodology determination, Jan

8 Chapter 1: Background information We will use our new customer-facing measures and targets to help prioritise our expenditure and to inform the measures, targets and revenue incentive regime we propose to the Commerce Commission. We will use the targets as the basis for aggregated annual targets that will be linked to our revenue we will be either rewarded or penalised for our overall performance against the aggregated targets. 1.5 How we will use our customer-facing performance targets The customer-facing performance targets are intended to provide individual customers with a clear idea of the long term performance they should expect at their grid connection. For example; the number of outages over a 10 or 20 year period. The targets will encourage us to focus spending where the benefits of improved service justify the costs. We will translate these customer-facing targets into internal measures to further refine our existing processes to reduce the risks and durations of outages for customers, improve asset performance and help us prioritise expenditure. Initially, the targets will be about delivering the service that the customer should expect given the combination of assets used (noting that the Grid Reliability Standards and customer investment decisions often dictate the assets employed to serve a connection). As the approach matures we will focus on matching the performance delivered to what is desired and economic. We don t expect we will spend large amounts adding more assets to achieve the performance targets. The targets should encourage us to find lower cost initiatives that deliver high benefit. The types of initiatives we embark on as a result of our initial targets will improve prioritisation of our capital and maintenance expenditure along with work to also improve asset or circuit performance, or restoration times. For example, at some parts of the grid it may be cost effective to have equipment on site to reduce the time to restore assets to service after an outage. We could install remotely operated cameras at substations or on lines to allow quicker restoration times following equipment failures and reduce the costs of calling out service personnel. Once we have exhausted lower cost options, further performance improvement at a grid connection will ultimately require expensive additions to major assets, but these are unlikely to be justified in isolation unless they are also required to enhance the capacity of the grid. Using customer-centric targets is not new in the New Zealand electricity industry. Table 1 gives examples of some of the customer focused targets currently used by distributors. Capital-Expenditure-IM/Capex-IM-Final-Determination-and-Reasons-Paper/Transpower- Capital-Expenditure-Input-Methodologies-Determination-2012.pdf 6

9 Chapter 2: What performance matters to customers? Table 1 Examples of distributor targets Service standards CBD Industrial Urban Rural Vector Wellington Electricity Orion Powerco Top Energy Alpine Energy Time to restore (hours) Number of Interruptions/annum Max time to restore (hours) % of fault restored in 3 hours Max number of interruptions Max average outage time Average number of interruptions Max outage duration (hours) Average number of outages % overall (3) (2) (2) (4-6) 2 What performance matters to customers? The grid transports electricity from generators directly to large industrial customers or to regional distributors who transport electricity to end consumers. The grid extends to the points at which our customers connect their assets to ours. Our customers typically either inject power into the system (generators) or take electricity from the grid (distribution companies and directly connected large consumers). We have looked at what we think the important performance issues are for off-take and injection customers and consumers those households and businesses supplied by distributors. Our focus is on the ability of the grid to transmit power without interruption. This is the core job of the grid and the area over which we have the most control. There are two other aspects of grid performance that our customers experience but we are not proposing to deal with here. Power quality (voltage sags and swells, harmonics, flicker and imbalances) is influenced by our investment and management of the grid. However, it is also influenced by the behaviour of customers connected to the grid. Power quality is largely managed by the system operator in response to the grid design, the 7

10 Chapter 2: What performance matters to customers? equipment connected to the grid and the real time activity on the power system. Additionally, the Electricity Industry Participation Code mandates certain power quality standards. Power losses across the grid and constraints in power transmission also depend on the market behaviour of electricity generators and retailers. The wholesale electricity market includes pricing mechanisms to take account of these and our investment planning processes consider these price effects. In addition, the Electricity Authority is implementing Financial Transmission Rights which will enable the market to manage the costs and risks associated with constraints. 2.1 Number and length of interruptions are important Our initial discussions with some distribution customers and directly connected large consumers (off-take customers) confirmed the main grid performance issues as the number and length of unexpected power outages. These outages happen infrequently but can cause significant disruption and cost. Most homes and businesses have no economic alternatives to reliable supply from transmission and distribution networks. Consumers are very reliant on their electricity supply as an essential service at work and home. Grid performance measures and associated targets need to reflect this dependence. For generators (injection customers), outages result in the unexpected loss of generation connection which can mean lost revenue or possible damage to equipment. However, generators sometimes do not have the same security requirements as distribution and directly connected consumers. This is reflected in choices which some generators have made to contract for a lower security grid connection. Both sets of customers want to minimise the cost of an unexpected outage. 2.2 Measures and targets should be meaningful to customers; system averages are of limited relevance Individual customers want to understand the reliability of their supply both the existing level and future expectations. This helps customers to plan how they manage outages, or whether to invest in alternatives such as back-up generation. Across New Zealand, we typically have around 140 unplanned power outages each year. This is an average of under 0.6 outages per year at each grid connection. But this average can be misleading as the chart below shows. 8

11 Number of outages per year Chapter 2: What performance matters to customers? Figure 1: Number of outages per year for individual customer grid connection points ( ) Average =0.6 0 Customer grid connections The chart shows that the majority of grid connections have less than 0.6 outages per year but the average is skewed by the performance at a small number of connections. For example, if we ignore the 10 worst performers, the average falls to 0.4 outages per year. Our existing performance measures rely on system averages which have limited meaning to customers. For example, two customers supply may fail five times each in a year, while their eight neighbours experience no outages. Alternatively, all ten could have one supply outage each in a year. In both of these scenarios, the average outage rate is the same. Nor do system average measures highlight where on the grid performance needs to be improved. They mask performance at the extreme limits; as illustrated above, average performance of 0.6 can mean superb performance for most customers and poor performance for a few. 2.3 Customer expectations vary Off-take and injection customers expectations can vary, and within these two customer groups, different customers have different needs and different expectations, depending on the needs of their consumers. For example consumers in rural areas might, generally, have lower expectations to those in a major city. Table 2 shows our view of the differing expectations. 9

12 Chapter 2: What performance matters to customers? Table 2: Differing expectations of supply reliability Load Type CBD Large industrial Small to medium businesses Reliance on electricity supply Critical for retailers High critical Backup generation Likely for larger retailers and office blocks Possible, but only partial High Not likely Few Expectations of outages Very few Very few Large city urban High Not likely Few Small city urban High Not likely Few Semi-rural High Not likely Few Rural Lower Possible for certain uses, such as dairies Some Back country Low Likely Many At most off-take grid connections, there will be a range of consumers with different expectations but it is generally not feasible to provide different levels of service. However many grid connections have a predominant load type ie largely rural or large city urban. Generators can also have varied service requirements depending on how costly it is to restart generators, what potential equipment damage might be incurred by an outage and power prices at the time of the outage. 2.4 Customers tolerance of events depends on the cause and how good we are at providing information Our customers and consumers have a greater acceptance of outages due to events such as earthquakes or snowstorms than those due to failures of grid equipment such as broken earth wires, or protection malfunctions. The provision of information on cause and likely length of an unplanned outage during the outage is also relevant to consumers. Information on duration can assist businesses that face decisions such as whether to shut a business for the rest of the day or for householders whether to attend to a freezer-full of food. 2.5 Overall performance matters; grid performance is only part of the overall service delivered to consumers For households or businesses, grid performance defines only part of the overall service received. Electricity supply depends on the performance of distribution networks also. There is little value in improving grid performance for a consumer if 10

13 SAIFI (interuptions per customer per year) Chapter 2: What performance matters to customers? that consumer s service is dominated by failures in the distribution network. Equally, any improvements in distribution performance should be matched by improvements in transmission performance. Transmission outages generally affect more consumers, and wider area outages have a disproportionately higher impact. This means the reliability standard for transmission are typically higher than for distribution. Outages due to transmission ideally should be a smaller proportion of overall outages. This is borne out in the Figure 2 below that shows transmission s contribution to the overall outages experienced by consumers over a four year period. Figure 2: Frequency of transmission and distribution outages for New Zealand SAIFI (distribution) SAIFI (transmission) 11

14 Chapter 3: Proposed measures and targets 3 Proposed measures and targets Based on what we think matters to customers and consumers, we are proposing: three different types of customer-facing performance measures; and to create a number of categories based on customers at grid connections and set different targets for the different categories. We have also included a fourth performance measure that compares transmission and distribution performance. This measure may be of value to customers or the Commerce Commission. For each of the measures we discuss the types of issues to consider in setting targets and include a table that shows illustrative targets. We expect to discuss ways to develop the targets in more detail at our meetings with customers and stakeholders. We will aim to progressively improve against the targets over a 10 or 20 year period. Although our performance against individual targets will not be directly linked to transmission charges, it will impact on the overall revenue we can receive from transmission customers. We will propose aggregate annual performance against targets (eg meeting 60% of our individual targets) during our next regulatory control period to the Commerce Commission. These will be the basis of a performance incentive regime where our overall annual revenue allowance is adjusted up or down based on our performance. 3.1 Different target categories one size does not fit all Different customers have different expectations and we think our targets should reflect this. A possible way to differentiate would be to set different targets for different grid connections based on the type of customers connected. It makes sense to look at the performance of off-take and generator grid connections separately Off-take grid connections Some distributors, for example Vector, Powerco and Top Energy, already differentiate their customer groups and specify different service standards for different types of customers (for example, urban, rural, CBD). Examples are shown earlier in Table 1. Our thinking is that it is appropriate to use a similar idea for performance at the grid connections, differentiating them into categories. The categories recognise the different criticality of the loads that they serve. The categories reflect the differing impacts of power outages, and also the increased costs that would be involved in meeting tighter targets for loads where outages lead to lower levels of economic and social impact. 12

15 Chapter 3: Proposed measures and targets Table 3 sets out our proposed categories for off-take grid connections. Table 3: Criticality of load categories Criticality of load Essential Important Standard Description of the load categories, with examples Grid connections that supply large CBDs or very large industrials. For example: Penrose (Auckland CBD), Addington (Christchurch CBD), Tiwai (aluminium smelter), Bream Bay (oil refinery). Grid connections supplying smaller cities or towns, supplying more than 25,000 ICPs or large industrials. For example: Bunnythorpe (Palmerston North), Kinleith (pulp and paper), Edendale (dairy). Grid connections supplying mainly rural areas. For example: Balclutha, Woodville. N-security grid connections Some grid connections are served by a single line or a single transformer and so any fault leads to a power outage. We propose that these N security connections will have different targets reflecting a combination of criticality and their increased exposure to events. Some distributors with N security have arrangements in place to keep the power flowing during outages through their ability to back feed from other grid connections or from generators within their own network. Although our objective during outages should be to return our assets to service as quickly as possible, an N security customer s ability to back feed may be relevant to possible targets Generator grid connections For generators, the targets will be focussed on loss of generation connection. As for off-take grid connections, the targets will reflect the security of their connections. 3.2 Number of power outages Proposed measure We propose to include all unexpected power outages at grid connections except the following: Outages less than one minute long. This is in line with international standards and New Zealand regulatory definitions 4. Most outages less than one minute long are known as momentary interruptions where power is generally automatically restored. 4 The Commerce Commission defines an interruption (referred to here as an outage) as cessation of electricity for a minute or longer in Commerce Commission Electricity Information Disclosure Requirements issued 31 March 2004 (Consolidating all amendments to 1 April 2007)). 13

16 Chapter 3: Proposed measures and targets Outages caused by faults on customer assets which are cleared correctly by our protection relays. At this point, we propose to include outages caused by severe environmental conditions (eg earthquake, snowstorm). We are interested in customer views on this. On one hand, it is hard to define these outages clearly enough to avoid extensive discussions around whether the conditions are or are not severe. On the other hand, when the extreme events are beyond economic design specifications these outages are beyond our control Targets The targets should reflect what performance should be expected at the grid connection for the different categories. At a grid connection built to provide N-1 security, losses of supply require a concurrent loss of two or more assets or systems. These losses could occur whilst one or more assets are out for maintenance, or when two unexpected outages happen at the same time. Our analysis indicates these sorts of outages should be expected around once every ten years based on typical maintenance plans and unplanned outage rates for assets. Common mode failures (protection settings, HEI s) or wider regional events suggest the return period for outages may be in the 2-10 year range. At grid connections with only N security it is often the length of the single line to the interconnected grid that determines the service, or the performance of a single transformer or other connection asset. We propose to set targets based on the individual circuits and the criticality of the load. We plan to express our targets in the form of a long term target (eg one outage every 10 years). We recognise we can only measure performance over a finite period (eg annually or over five years). For this reason we will set targets for a maximum number of outages in 1 year and over 5 years. If the long term target (for period of 10 or 20 years) at the grid connection is one outage every five years and we experience 0, 1 or 2 faults there in a five year period then it is statistically likely that we are meeting the target. Table 4 below shows the probability of observing a particular number of outages in a five year period if we have a long term outage rate of one every five years. 14

17 Chapter 3: Proposed measures and targets Table 4: Probability of faults Number of faults over 5 years Probability of observing that number of faults (assuming a Poisson distribution) 0 37% 1 37% 2 18% 3 6% 4 or more 2% Number of power outages: what our targets might look like Table 5: Example of performance targets for number of outages Category Long term average outage rate Maximum number of outages in 1 year Maximum number over 5 years Essential Important Standard Generator We are still developing targets for N-security sites, but expect to base them on a combination of criticality and exposure to events. 3.3 Duration of power outages Proposed measure We propose to record the length of all outages greater than one minute long, originating in our network, including those caused by extreme events such as earthquakes and storms. We will measure up to the point when we begin to restore supply. It is not clear whether we should measure duration up to the point where a customer back feeds, or is able to load manage. We are interested in customer views on these issues Targets We restored power following outages within 30 minutes for more than half of our outages in the last 10 years, and took longer than 8 hours for less than 1 per cent. In short, although a large amount of events are dealt with relatively quickly, there are some significantly long outages for some customers. Our proposal is to set targets based on average duration of outages at each grid connection, but also to set targets for the length of the longest outages. The targets will differ by category. We also propose to set different targets for restoration times after major environmental events such as earthquakes or storms. By their nature, these events present significant challenges for supply restoration. 15

18 Chapter 3: Proposed measures and targets Duration of power outages: what our targets might look like Table 6: Example of performance targets for outage duration Category Average duration (min) 90 th percentile duration (min) Essential Important Standard Generator Major environmental event (earthquake, storm etc.) Again, we are still developing targets for N security grid connections. 3.4 Information provision during power outages Proposed measures Once an outage occurs, accurate information about restoration is valuable to customers. We currently communicate with our customers (distributors, large industrials and generators) during an outage. Depending on the extent and length of the power outage we also provide public information via our website and the media. Presently we do not communicate with households and businesses, although we could communicate through social media such as facebook and twitter. Possible performance measures are: Time to make initial contact with the distributor, generator and or major industrial customers. Time to communicate expected restoration time, and the accuracy of updates compared to actual restoration time Targets Potential targets are to: inform the nominated customer within 15 minutes of an outage; and communicate an initial assessment and time to restore within 30 minutes of the outage and provide hourly updates. 16

19 Chapter 3: Proposed measures and targets Information provision: what our targets might look like Table 7: Example of performance targets for information provision Measure Time to inform customer Time to communicate expected restoration time Accuracy of estimated restoration time compared to actual restoration time Target 15 min 30 min +/- 15 min 3.5 Transmission outages relative to distribution Proposed measure Our performance is already compared with international transmission companies, and distribution company performance is compared within New Zealand. These comparisons provide useful benchmarking information. A measure that compares transmission outages to distribution in New Zealand might be valuable to customers and a useful measure for us to monitor and report in order to understand transmission s part in the overall performance that customers experience. We could measure either a comparison of the number of outages or the outage duration. Over the four years from 2008 to 2011, the ratios have been similar about 24 per cent of both the number of outages and the total length of outages have been due to transmission. Unlike the first three measures, this measure uses system-wide averages. We propose this measure as a benchmark to provide customers and the regulator with information Targets We could benchmark our performance against the collective performance of the lines companies on a rolling 10 year average, possibly for both number of outages and duration of outage. An appropriate target might be for outage duration due to transmission to be no more than 20 per cent of all outages experienced by consumers. We do not propose to link this measure to revenue in our proposal to the Commerce Commission for our next regulatory control period Transmission compared to distribution: what targets might look like Table 8: Example of performance targets for transmission and distribution comparisons Measure Target Proportion of all outages due to transmission 20 % Proportion of outage duration due to transmission 20% 17

20 Chapter 4: Your feedback and next steps 4 Your feedback and next steps This paper has set out our thinking on appropriate customer-facing grid performance measures and targets. We would like to hear what our customers think before finalising the measures and targets, using them to guide our planning, and incorporating them into our RCP2 proposal to the Commerce Commission. We include a feedback form (Appendix A.1) which you can use. We will also be meeting with customers to both discuss the ideas in more detail. So we can consider your feedback we need to receive it by 5 December Please send your feedback to: Graeme Ancell, Power System Planning Manager, Transpower at: PMFeedback@transpower.co.nz. 18

21 Appendix A.1: Feedback form A.1 Feedback form Please add your feedback to this form. You are welcome to provide additional feedback. Questions 1. What matters to customers (Section 2) Our preliminary views are that the following are important issues to consider in redeveloping our performance measures: The number and duration of outages are the most important factor for customers. Measures and targets should be meaningful to individual customers; system averages are of limited relevance to customers. Customer expectations vary. Customers' tolerance of an outage depends on the cause and how good we are at providing information. Overall performance matters; grid performance is only part of the service delivered to consumers. Do you agree with these? (yes/no) What other matters are important to you as a Transpower customer, or do you think are important to consumers? 2. Different target categories (Section 3.1) Do you think it a good idea to set different targets for different categories of customers at the grid connection? (yes/no) Do you think our proposed categories make sense or do you have any other suggestions? 19

22 Appendix A.1: Feedback form Questions 3: Number of power outages measure. Section 3.2 Do you agree with the definition of the number of power outages measure including the proposed exclusions and the way we propose to set the targets? (yes/no) Do you have any comments or suggestions? Do you think we should include or exclude outages caused by extreme conditions such as earthquake or snow storm? 4: Duration of power outages Section 3.3 Do you agree with the definition of the duration of power outages measure including the proposed exclusions and the way we propose to set the targets? (yes/no) Do you have any comments or suggestions? What are your views on how we should measure durations where customers backfeed or load manage during a transmission problem. 5: Information provision Section 3.4 Do you think the information provision measures we propose will be useful? (yes/no) Do you have any comments or suggestions on the proposed measures and targets? 20

23 Appendix A.1: Feedback form Questions 6: Transmission outages compared to distribution Section 3.5 Do you think a measure to compare transmission and distribution performance is useful? (yes/no) If so, which of the two measures is the most important? 21