# Existing plants The base year generation data is intended to reflect the distribution of 2023 generation across all generating New Zealand assets. These should be available to the model to meet future demand, but with enough information (region, technology, remaining life, etc.) that the model will retire plants at appropriate points and can make least-cost dispatch and peak decisions. TIMES-NZ requires detailed information on the existing generation stock, including: - Capacity (GW) - Fuel inputs and efficiencies - Maintenance costs, including fuel delivery costs. All costs are in 2023 NZD - Asset lifetime and current age - Region (North Island/South Island) - Peak contribution availability - Capacity factors (the upper limit on the proportion of the year the plant spends generating at full output) - Generation levels in the base year, to calibrate historical data (GWh) - Availability curves for intermittent generators, such as wind or solar We have improved on TIMES-NZ 2.0 by building a bottom-up, asset-based model of the existing generation fleet. This gives us much greater detail in how the model utilises the existing generation fleet and will allow us to make very precise changes as needed. ## Existing asset list EECA has prepared a list of current plants for the base year generating stock[^eeca_plant_list]. It is closely based on the EA dispatch generation fleet list[^ea_dispatch_generationfleet] and includes mapping to the current plants from MBIE’s EDGS Generation Stack[^edgs_assumptions]. Where possible, we have included mapping to plant names found in the Electricity Authority’s “Generation_MD” estimates of generation by plant[^ea_md_generation]. The asset list has been reviewed and updated based on developer statements and Energy News[^energy_news] resource files for each plant. In some cases, capacities were updated, or names were adjusted slightly. This list is not intended to capture all distributed or cogeneration facilities, which are instead represented by generic plants (see below). In general, plant status has been set to align with MBIE’s EDGS generation stack categories for 2023. This allows us to also use the MBIE generation stack for potential future generation assets without double-counting. For example, Kaiwera Downs Stage 1 is included in the base year, but the rest of the build is considered a future technology. This is because Kaiwera Downs was only partially operational by the end of 2023. [^energy_news]: Energy News: [^ea_md_generation]: Generation by plant | EA: [^eeca_plant_list]: The current version of this plant list can be viewed on our github [here](https://github.com/EECA-NZ/TIMES-NZ-Model-Files/blob/main/PREPARE-TIMES-NZ/data_raw/coded_assumptions/electricity_generation/GenerationFleet.csv). [^ea_dispatch_generationfleet]: Existing Generation Plants | EA: [^edgs_assumptions]: Electricity Demand and Generation Scenarios - Assumptions Data | MBIE: ## Distributing base year generation using Electricity Authority data We use the Electricity Authority’s “Generation_MD”[^ea_md_generation] data to find estimates of generation for the current plant list. This bottom-up approach means we can assign known generation to plants, regions, and technologies. This data covers 93%[^93_percent] of total generation for 2023. For the remaining generation, we make some assumptions on the location and technology of distributed generation (cogeneration or otherwise) to calibrate final figures with historical generation and capacity data published by MBIE[^mbie_official_generation]. For plants where we include multiple stages, but there is only one reference in the Electricity Authority data (such as Ngāwhā or Turitea), generation is proportionally distributed by the stage’s capacity. We further test our final base year generation figures against MBIE’s official historical statistics to ensure accuracy. [^93_percent]: 40,597 GWh in the EA’s dataset is 93% of the 43,494 GWh reported by MBIE for 2023. [^mbie_official_generation]: Electricity statistics | Ministry of Business, Innovation & Employment: ## Capacity factor estimates In some cases, using generation data from the Electricity Authority is either incomplete or results in implied capacity factors outside normal ranges. This is especially true for some cogeneration plants but could also indicate that our manual asset mapping to Electricity Authority plant definitions was inaccurate. In these cases, we instead set some plants to have generation estimated by capacity factor. These capacity factor assumptions are used for calibrating the base year and provide upper limits to future generation for these plant types. These capacity factor assumptions are either extracted from TIMES-NZ 2.0, or (where these were not available) estimated based on 2023 base year calibration. More detail on capacity factors and availability factors for plants are available in the [Availability Curves](availability_curves.md) documentation. ## Huntly Rankine units To estimate the split of coal and natural gas at the Huntly Rankine units, we assume the Rankine 2023 generation from coal is equivalent to MBIE figures on total coal electricity generation (i.e. excluding cogeneration), and the remaining share is from natural gas. The resulting figures align with Genesis reporting on Huntly coal use during calendar year 2023[^huntly_coal_use]. ```{eval-rst} .. admonition:: A note on biomass :class: note In February 2023, Genesis completed a biomass trial at the Rankine units. This means some small proportion of the generation was fuelled by imported wood pellets. These figures are negligible overall, and they are currently not included in the base year model. ``` [^huntly_coal_use]: Calendar values of coal use at the Rankine Units can be found by compiling quarterly report values from ## Distributed solar generation The Electricity Authority does not provide figures on rooftop solar generation. However, distributed solar capacity data is available. We therefore create generic plant stocks intended to represent different levels of rooftop solar generation (residential, commercial, and industrial), and distribute MBIE’s official solar generation statistics according to region and island based on EMI distributed solar capacity data. A stock model is applied to the existing stock of distributed solar generation to estimate the rate at which panels are retired from rooftops across the model horizon. Plants linearly retire according to the technical lifetime assumption for solar plants. ## Adding generic plants After adding either Electricity Authority generation data or plant generation estimates based on capacity factors, we calibrate total figures against official MBIE generation data (broken down by cogeneration status and fuel type). We expect to still be missing some generation when comparing to official statistics, which reflects smaller embedded or other plants not available in our plant list or Electricity Authority solar capacity data. We therefore add several “generic” existing plants. These have their capacities and generation figures for the base year generated automatically based on the missing generation data and capacity factor assumptions. In cases where they may be on either island (such as wind or hydro), they are distributed according to the known regional distribution of similar plants. In other cases (such as geothermal or natural gas plants) they are distributed only across the North Island. ## Technical parameters from MBIE data Any generating asset in the model requires detailed technical parameters. The following are available in the MBIE generation stack for current plants: 1) Heatrate (or fuel efficiency) 2) Variable O&M (NZD/MWh) 3) Fixed O&M (NZD/kW/year) 4) Fuel delivery costs (NZD/GJ) In many cases, we can directly map base year generating assets to MBIE plant names to add these parameters. In cases where plants in the TIMES-NZ base year do not have direct equivalents, we take the mean of these parameters for similar plants and apply those instead. For geothermal, we do not directly use the default heat rate implied by the MBIE generation stack. If applied as-is, this would understate geothermal fuel input relative to MBIE's official energy statistics. We therefore override existing geothermal generation to use a heat rate of 24,000 GJ/GWh, equivalent to an electrical efficiency of 15%. ## Technical parameters by assumption Other required parameters are not included in MBIE’s data, so these are applied by assumption. ### Peak contribution rate The peak contribution is the assumed proportion of capacity available to meet peak demand. These peak contribution rate assumptions are as follows: ```{list-table} Peak contribution assumptions :header-rows: 1 :name: tab-peak_contribution * - Plant Type - Peak contribution (%) * - Hydro (dispatchable) - 90 * - Hydro run-of-river - 72 * - Peakers (all fuels) - 97 * - Natural gas (CCGT) - 97 * - Wind (onshore) - 10 * - Wind (offshore) - 25 * - Solar - 0 * - Huntly Rankine Units - 97 * - Geothermal - 92 * - Biogas - 66 * - Battery storage - 98 * - All other generation - 66 ``` For intermittent generation, such as wind and solar, peak contribution rates are based on the lower likelihoods of availability during peak hours (winter 6pm-7pm). Onshore wind peak contribution rates are based on the 10th percentile of historical wind availability during peak hours (winter 6pm-7pm). This figure comes to around 10%. Offshore wind peak availability is by assumption, but we expect that offshore wind resource will be more reliable than onshore. We assume solar is not available when it is dark. Dispatchable hydro peak contribution rates are 990%. This accounts for consenting and reserve capacity requirements and the possibility of reduced hydro availability during peak periods. Other peak contribution rates are assumptions extracted from previous iterations of TIMES-NZ. ### Plant lifetimes All generating assets are provided with their original commissioning date where possible. The technical lifetimes are by assumption for each technology type. If a plant reaches the end of its technical lifetime, it is decommissioned in the model. These assumptions have been extracted from TIMES-NZ 2.0. ```{list-table} Plant lifetime assumptions :header-rows: 1 :name: tab-plant_lifetimes * - Plant Type - Technical lifetime (years) * - Biogas - 30 * - Generic cogeneration - 25 * - Diesel (peakers) - 30 * - Geothermal - 60 * - Hydro - N/A * - Huntly Rankine units - 50[^huntly_life] * - Natural gas - 25 * - Solar (distributed and utility) - 20 * - Wind (onshore and offshore) - 25 ``` Currently, decommissioning costs are not included. This was also true in TIMES-NZ 2.0. We instead assume that plants always retire at the end of their technical lifetime without incurring decommissioning costs. We also assume hydro plants are maintained through the entire model horizon; hence the technical lifetimes are `N/A`. We assume the cost of turbine replacements are spread across their operating and maintenance costs each year. ### Lifetime adjustments The above lifetime assumptions apply broadly to all technology types. However, we make a few manual adjustments to some plants to force specific closure dates when these don't align with the technology lifetime assumption. These are as follows: ```{csv-table} Plant lifetime adjustments :header-rows: 1 :name: tab_plant_lifetime_adjustments Plant, Adjusted closure year Te Rapa, 2024 Wairākei, 2027 Whareroa, 2030[^whareroa] ``` All other plants will retire when their age exceeds their technical lifetime based on initial commissioning date. Note that we have also extended the lifetime of the Tararua wind farm to 30 years based on observed reliability. [^whareroa]: The Whareroa plant at Hāwera has had its lifetime extended, but 2030 is assumption only. [^huntly_life]:The Rankine units were commissioned across 1982-1985, and we have set the commissioning date to 1983. This lifetime assumption means they will be retired in the model in 2033.