National Electricity Market Management Company Limited ABN 94 072 010 327 Interval Meter Rollout Prepared by: Metering & Retail Development 28 February 2003 Reference: ESC Victoria, November 2002 Installing Interval Meters For Electricity Customers Costs And Benefits 1 This paper provides NEMMCO s comments on proposals outlined in the above paper. The comments only address implications for NEMMCO s Market Settlement and Transfer Solution (MSATS) system and related business processes. Comments are not provided on costs or benefits to industry, or implications for NEM participant business processes to support an interval meter rollout. 1. General The National Electricity Code already provides for the use of interval meters in the NEM and NEMMCO s MSATS processes and procedures already support the application of interval meters in retail market processes and NEM settlement. The current MSATS system design, as agreed by NEM participants and jurisdictions, supports the processing of data from a maximum of 552,000 interval meters across the NEM. This has been verified in benchmark tests. As detailed below, while there are no benchmarks available that confirm whether a greater volume of interval metering data can be processed adequately or whether functional limitations exist in the MSATS implementation, the design is intended to scale linearly beyond 552,000 interval meters and NEMMCO is not 1 http://www.esc.vic.gov.au/attachmentviewer1058.html
aware of any inherent limitations. While it may be possible to handle the initial years of a rollout strategy by upgrading the MSATS equipment, benchmark analysis will be required before we can confirm whether the current MSATS implementation will handle a very large volume of interval metering data. The cost / benefit analysis included in the ESC, Victoria (the Commission s) paper does not appear to include NEMMCO system costs to process the additional interval metering data. As information on these potential costs was not sought from NEMMCO, it is assumed that the cost / benefit analysis considers the cost per meter to be small when compared to meter installation and data collection costs and, therefore, not significant to the cost / benefit analysis. While this may be adequate for the analysis, it will be necessary at an appropriate time to identify these costs and consider how they should be recovered. NEMMCO s costs would initially continue to be recovered through participant fees from retailers, which are proposed to be based on energy transactions and not meter types or numbers. If an interval meter rollout in one jurisdiction were to result in significant costs to process a large volume of metering data, it may be necessary to consider whether these costs should be recovered only from retailers with customers in that jurisdiction. However, once NEMMCO s fee structure for the three years from July 2003 has been determined, it will not be possible to change the structure until the next fee determination, for the period starting in July 2006. 2. Current MSATS Design The MSATS design includes a central CATS (Consumer Administration and Transfer Solution) server that manages metering standing data and synchronises changes to that data, including retrospective changes, with satellite MDM (Metering Data Management) servers that calculate and apply profiles and aggregate energy consumption data. The design currently provides for the following: Management of standing data from 10 million metering points, with daily synchronisation of MDM processes. Aggregation of up to 10 million metering data streams (energy consumption data), including up to 552,000 interval metering data streams and the processing of profiles for accumulation metering data, for NEM settlement purposes. Two current MDM servers handle the following metering data stream volumes: MDM 1 = 21,269 interval + 119,352 basic (accumulation) metering data streams. MDM 2 = 21,989 interval + 1,479,507 basic (accumulation) metering data streams. Testing has confirmed the following: The main CATS server can handle the projected 10 million metering points. It is anticipated that a greater volume could be managed, if required, depending on the volume of consumer transfers and metering configuration changes also involved, but could require the system hardware processing capability to be increased. 28/02/2003 2
Each MDM server can handle up to 72,000 interval metering data streams and more than 2 million basic (accumulation) metering data streams. MDM servers can be added as required, with no reduction in the capacity of each MDM server and no significant increase to total MSATS processing times. However, as the number of interval meters increases in each MDM server there may be a need to consider modifications to the process, such as moving each profile area into a separate MDM server, or creating progressively smaller profile areas (i.e. fewer transmission nodes in each area) with each area running in a separate MDM server, or separating the processing of interval metering data and basic (accumulation) metering data. 3. Projected Metering Data Volumes in Victoria The Commission s paper indicates the following metering data volumes: 3.1 There are currently 2,184,810 consumers with basic meters It should be noted that this translates into 2,854,737 basic meters in Victoria in the MSATS database (i.e. an average of 1.3 meters per consumer). It is not clear whether the Commission s strategy would replace multiple basic meters with single interval meters, which would require time-of-use tariffs, however the totals below assume one interval meter per consumer. If meters were replaced on a one-for-one basis, the volume of metering data could be 30% greater. 3.2 New & Replacement Strategy (interval meters to be installed): New consumers per year = 1.2 % (e.g. = 26,220 p.a.) Replacements per year of non-off-peak meters = 3% (e.g. = 47,800 p.a.) (off-peak replacements included below) 3.3 Rollout Strategy: Large consumers (>160 MWh/y) = all meters, over 2 years (e.g. = 2,000 p.a.) Small consumers, 3-phase = all meters, over 5 years (e.g. = 41,818 p.a.) Small consumers, off-peak = all meters, over 5 years (e.g. = 118,303 p.a.) 3.4 Totals: Years 1 & 2 = 236,141 p.a. Years 3, 4 & 5 = 234,141 p.a. Years 6, onward = 74,020 p.a. Total after 2 years = 470,000 Total after 5 years = 1.2M Total after 10 years = 1.5M 4. Projected Metering Data Volumes in Other Jurisdictions SA and NSW have indicated that they may also consider the rollout of interval meters and if this happens the ACT may also follow. Qld is yet to proceed with FRC and any interval meter rollout is likely to be linked to their FRC plans. Tasmania anticipates proceeding with FRC a few years after they enter the NEM. The MSATS design should 28/02/2003 3
therefore provide for an interval meter rollout in all jurisdictions, including Qld and Tasmania. Assuming that the number of metering points in Victoria is approximately 33% of the NEM and anticipating that all jurisdictions may rollout interval meters, the following is anticipated: The volume of interval metering data after 5 years = 3.5 million data streams The volume of interval metering data after 10 years = 4.6 million data streams 5. MSATS Design Implications 5.1 Core Design The proposed interval meter rollout doesn t change the need to synchronise all consumer transfers and all changes in metering details, particularly retrospective changes, with the processing of energy consumption data for NEM settlement. The proposal will also not alter the need for MSATS to calculate and apply profiles to accumulation metering data, at least not for several years. The volume of accumulation meters requiring profiles would progressively reduce, but even after ten years of the proposed interval meter rollout there would still be more than 680,000 accumulation meters that require profiles in Victoria (and possibly 2M across the NEM). Hence the core MSATS design is still required. 5.2 MDM Design Metering data associated with Victoria s five profile areas and the SA profile area is currently processed one MDM server. As the volume of interval metering data increases it is possible to modify these MDM arrangements. Alternatives that can be explored include the following: Increase the capability of the MDM process by expanding the current servers, or purchasing newer servers and/or improving the application design. Server performance is currently doubling every 18 months, which means that new servers should be able to process twice the volume of data processed by current servers. Move each profile area into a separate MDM server, possibly using higher performance servers that can process the full volume of interval metering data anticipated in each profile area. Redefine the profile areas to reduce the number of transmission nodes and, hence, metering points in each area and spread the profile areas across several MDM servers. Modify the MDM process design to allow the processing of interval metering data and profile calculations in separate servers. However, there are several related implementation issues that will need to be addressed before an appropriate solution can be identified. For example, the reporting of interval metering data to participants for settlement reconciliation purposes is a major processing load for NEMMCO and participant systems. The additional volume of interval metering data may, therefore, require improvements in the reporting process. The processes that 28/02/2003 4
load metering data into MSATS, including NEM network capabilities, may also need to be reviewed. While detailed design work still needs to be done, we believe that solutions can be found to match the volume and rate of interval meter rollout at moderate cost. It isn t possible to accurately estimate costs until this detailed analysis is undertaken. However, our initial analysis indicates that equipment to process the 1.2 million interval metering data streams anticipated in the first 5 years of the rollout strategy would cost $400,000 over 5 years (budget estimate, only), plus operating costs. There would also be a moderate additional cost to increase the capacity of the NEM network to handle the increased data volume. 5.3 External metering service providers As part of the analysis of alternative MDM arrangements it may also be appropriate to review the role of metering service providers in the processing of interval metering data. It is assumed that these service providers would collect and process this data for retail and network purposes and would supply the data to NEMMCO for NEM settlement purposes. Our initial analysis indicates that it is likely to cost less to upgrade the current MSATS process, as above, than to significantly change the MSATS design. The current design requires all interval metering data to be delivered to NEMMCO for use in profile (net-off) calculations and for synchronising NEM settlement aggregation with consumer transfers and changes to metering details. However, the service providers may need to increase the size of their data connections to the NEM. 6. Next Steps The Commission s paper advises that comments provided in response to its November 2002 position paper will be used to develop a draft decision on the proposed interval meter rollout and that further discussions and round-tables are likely to be held to address issues. Proposals would then be sought from distributors for the rollout of interval meters, which would be assessed prior to making the final decision. NEMMCO s MSATS strategy for interval meters also needs further clarification and would depend on the proposed rollout programmes of distributors. It would be appropriate if the Commission s considerations included time for NEMMCO to clarify the MSATS requirements once the distributors proposals are known, before making a final decision on the interval meter rollout programme. Before an initial MSATS strategy can be confirmed, NEMMCO needs to establish performance benchmarks for the application design, MDM systems, data storage requirements and associated operational issues. This work would take three months and cost $100,000 (est.) and could be undertaken within the next 6 months. Subsequent design work would be needed before MSATS strategies can be confirmed for the volume of interval metering data anticipated in 5 to 10 years. When developing the rollout programme, 12 months needs to be allowed after the final decision is made for the existing MDM servers to be upgraded to accommodate the interval metering data volume anticipated in the first two years of the programme. 28/02/2003 5