Retail Electricity Prices

[Ethan-Russell]

Need to add an easy way to compute retail electricity prices for a region. This is a cross-table calculation that uses many of the welfare terms. The total retail electricity payment equals:

• + electricity cost from LMP
• + generation standard payment to producers
• + reserve payments
• + distribution cost
• - merchandising surplus
• - cost of service rebate

[Ethan-Russell]

I'm thinking of making a little adjustment to the welfare structure. It's great that we can track down individual terms with the current form. I think it would be good to make it more similar to the way results are being done now.

[mckennapep]

A retail price mod has now been set up that calculates retail prices similar to the structure that Ethan described above. It is a cross-table calculation that is set up similar to the welfare calculation. A RetailPrice mod was created with much of the modify_results function following the set up in the ResultsTemplate mod.

In data, the setup_retail_price! function from the newly created retail_price.jl script is called which grabs the following terms from multiple tables and stores the terms in the retail_price table in data:

• + electricity cost from LMP
• + generation standard payment to producers
• + reserve payments
• + distribution cost
• - merchandising surplus
• - cost of service rebate
• + cost of past invest costs

Notes:
The cost of service rebate term is included to account for the costs of imports/revenues from imports in cos regions. This term includes electricity revenue, production cost, and net policy costs. When electricity revenue is greater than electricity costs, the production costs should be lowered so that rate payers aren't bearing the costs of producing exported energy (and vice versa when costs are greater than revenue).

There was already a way to calculate past investment costs in e4st but it wouldn't be as accurate at a state level (or any aggregated version). This is because the gens are aggregated and past investment cost calculation relies on the age of the generator, which would be an average across the generators. For this reason, an option was added to calculate past investment costs using a full gen file that is read in. When this file is included, it overwrites the calculation from the gen file and is included as a separate term in the retail price calculation.

When the RetailPrice mod is called it will compute the retail price for the region, year combination specified in the config file. The config file will also specify the cal_mode which is used for calibrating the calculated prices with reference prices. When set to none, the calculation will be done as usual. When set to get_cal_values a reference price file will be read in that is used to find the price adders. When set to calibrate a set of price adders will be read in that are added to the calculated retail rates. There is an option to provide 1 year of historical prices that will be used as the reference price, or provide annual price that are projected into the future. When 1 year is provided, a single calibrator value for each region will be found that is then applied to every sim year. When multiple years are provided, a different calibration value is found for each region/year combination. When a historical year is used, inputs should closely match ground truth for that year (i.e. gen set, demand, policies). To use yearly reference prices, it is important to set the model up in a similar way to the modeling that was used to get the price projections (i.e. if the prices are from aeo, the calibration run should use AEO demand, price, policy inputs). Currently, we use state level prices from EIA for the year 2024 to get calibration values for each state. The reference prices are stored in e4st-input-processing/data/config/mods/retail-price-calibration

A second merchandising surplus formula was included that only finds merchandising surplus for competitive regions because congestion rent is not distributed back to rate payers in cos regions.

[mckennapep]

Done in #348