CAISO Methodology

Modo Energy provides benchmark data for battery energy storage systems across global energy markets, applying a standardized methodology to ensure consistency and transparency across all produced Indices.

1. Introduction and rationale

The Modo Energy CAISO Virtual Asset Methodology sets out how Modo Energy's virtual battery energy storage benchmarks for the California Independent System Operator (CAISO) are constructed. It explains:

• what the Benchmarks measure and why this method is used (§1);
• the key terms used (§2);
• the representative asset and how values are calculated (§3);
• the revenue components included (§4);
• the data inputs, fallbacks and minimum data requirements (§5);
• the dispatch model (§6);
• how expert judgment is exercised (§7);
• how the Benchmarks are determined when data is missing or markets are disrupted (§8);
• when published values may be revised (§9);
• how the methodology is reviewed and changed (§10–11);
• and the Benchmarks' limitations (§12).

1.1 What the CAISO virtual asset Benchmarks measure

The ME vBESS CAISO Benchmarks measure the simulated market revenue of a representative grid-scale lithium-ion BESS in CAISO, expressed per MW of rated power. The family of benchmarks is grouped by pricing location so revenues can be compared across California.

Each Benchmark reflects revenue a representative asset could earn from CAISO's energy and ancillary service markets rather than a measurement of any individual operator.

1.2 Why CAISO requires a simulated asset-level benchmark

The CAISO virtual asset Benchmark is simulated, not measured.

CAISO does not publish the asset-level data needed to account for the operations and revenues of individual batteries on its grid. It publishes market clearing prices for energy (locational marginal prices, LMPs) and for each ancillary service, but not per-asset dispatch, awards or revenue. By contrast, ERCOT publishes the ancillary service awards and telemetered net output of individual units, which allows Modo Energy to construct measured benchmarks for ERCOT.

Unlike other ISOs in the United States, CAISO does report the operations of its battery fleet as a whole. Its Daily Energy Storage Reports publish fleet ancillary service capacity awards and energy schedules for each market run. This data is used to construct the Modo Energy CAISO System Benchmark, and in this methodology to limit how much ancillary capacity the representative asset can clear (§5.4).

1.3 Why this method is used

A fleet-level benchmark shows what CAISO's batteries earned in aggregate. It cannot show what a single, standardized asset would have earned at a specific location. Modo Energy therefore simulates a representative battery trading against the prices CAISO published, using its Global Dispatch Model (GDM): a mixed-integer linear program (MILP) that maximizes battery revenue subject to market rules and physical constraints (§6). This approach:

• uses public CAISO data, so every input can be traced to a CAISO publication or to Modo Energy's CAISO asset database;
• applies a fixed, disclosed asset specification at each location, so changes in Benchmark values reflect changes in market prices rather than changes in the asset or its operator; and
• uses the same model and structure as Modo Energy's simulated benchmarks in other markets, so CAISO values can be compared with them.

Because the model optimizes with perfect knowledge of the day's prices, its raw output overstates what a real operator would earn. A calibration factor is applied to account for this (§6.6).

2. Definitions

3. Benchmark Construction

This section details the parameters that define the simulated battery in a Modo Energy virtual asset benchmark. The default values for those parameters in the ME vBESS CAISO Benchmarks are also provided.

3.1 The ME vBESS CAISO Benchmarks

Modo Energy produces a family of six benchmarks for CAISO. The benchmarks vary in the aggregated pricing node of the simulated battery:

• There are three load aggregation points, representing the three largest load serving entities in CAISO: SCE (DLAP_SCE-APND), SDGE (DLAP_SDGE-APND), PGAE (DLAP_PGAE-APND)
• There are three trading hubs: SP15 (TH_SP15_GEN-APND), NP15 (TH_NP15_GEN-APND), ZP26 (TH_ZP26_GEN-APND)

The naming convention for this family of benchmarks is ME vBESS CAISO [pricing node] ([duration]H) — for example, the 4-hour benchmark for the SP15 hub is "ME vBESS CAISO SP15 (4H)".

This family of benchmarks are produced using the methodology described in this document, with a common set of operational and behavioral parameters. These parameters are specified below.

3.2 Representative vBESS specification

These parameters are set by Modo Energy, and are under the purview of the US modeling team. They define the representative asset and the behavior of the dispatch model. No input data (prices) are modified.

All parameters listed in §3.1 and §3.2 are customizable in user-defined vBESS benchmarks, with the exception of AS expected throughput and calibration factor.

3.3 How the Benchmark is calculated

For each operating day, the Benchmark is calculated by:

1. Running the GDM for the operating day using the representative asset (§3.2) and the input data (§5).
2. Computing simulated revenue in each modeled market, in nominal USD, as the cleared price multiplied by the simulated position in each interval, plus the settlement of ancillary deployment energy at the RTD price (§4.1).
3. Multiplying energy and ancillary revenue by the calibration factor of 90% (§6.6).
4. Summing revenue across markets and dividing by rated power (100 MW) to give a value in USD/MW for the operating day.

Values for longer periods (for example, a month) are the sum of the daily values in that period.

3.4 Benchmark value representation

All Benchmark values are reported as revenue per unit of rated power, in the following units:

• USD/MW (period): total Benchmark revenue for the period (e.g. USD/MW/month).
• USD/MW/year (annualized): period revenue divided by the number of days in the period and multiplied by 365.

3.5 Publication cadence

The Benchmarks are updated daily. Modo Energy ingests CAISO day-ahead and real-time energy and ancillary service prices, together with the Daily Energy Storage Report, and publishes for the most recent operating day for which the minimum data in §5.5 is available, after a short lag to allow for CAISO settlement and late postings.

3.6 Benchmark coverage

4. Revenue Components

Modo Energy's virtual asset BESS benchmarks for CAISO only consider wholesale revenue opportunities: energy and ancillary service. Resource Adequacy contracts, PPAs, and other capacity-provision like agreements are not included.

4.1 Revenue components included in the Benchmarks

All ancillary products are capacity products: the battery is paid per MW of reserve held. Regulation Up and Regulation Down are separate products, each with its own clearing price. Energy delivered or absorbed when an ancillary service is deployed moves the battery's state of charge, counts toward cycling, and is settled at the RTD LMP, as CAISO settles it (§6.5). That settlement is reported within RTD energy revenue.

4.2 Excluded revenues and costs

• Regulation mileage payments: paid per MW of movement rather than per MW of capacity; not modeled.
• Resource adequacy (RA) payments: bilateral and not continuously cleared or publicly priced; not included in the Benchmark.
• Congestion revenue rights / congestion hedging: portfolio-specific; not modeled.
• Curtailment compensation: site-specific; not modeled.
• Network / transmission access charges: site-specific; set to zero.
• Bilateral contracts (PPAs, tolls, floors): not modeled.
• Operator fees, warranty, O&M: the Benchmark reports gross market revenue, not net of operating costs.

5. Data Inputs

5.1 Inputs and sources

Prices are collected from CAISO's OASIS interface:

Three further inputs describe how ancillary awards behave in practice:

5.2 Pricing locations and AS regions

Energy is priced at the LMP of each benchmark's pricing location (§3.1). All benchmarks price ancillary services in the AS_CAISO_EXP region. Of the AS regions that appear in CAISO's ancillary price data, AS_CAISO_EXP is the only one with consistently non-zero prices, and it matches the region used by the ME BESS CAISO System Benchmark.

5.3 How inputs are combined

• All inputs are aligned to a common 5-minute grid in Pacific Prevailing Time. Hourly and 15-minute prices apply to each 5-minute interval within their period.
• Energy and ancillary prices are used as published by CAISO. Modo Energy does not cap, filter or adjust prices, including during scarcity events.
• Acceptance rates set an upper limit on ancillary volumes (§6.3); expected throughput sets the SOC effect of ancillary positions (§6.5).
• Revenue from each market is computed separately and then summed (§3.3).

5.4 Acceptance rates and expected throughput

Acceptance rates. CAISO's battery fleet is many times larger than its total ancillary requirement, so nearly all ancillary awards go to batteries, and the share any single battery can win falls as the fleet grows. The model caps the capacity the representative asset can hold in each service, in each interval, at its rated power multiplied by an acceptance rate. The acceptance rate is the fleet's awarded MW in that service divided by total fleet rated power:

• Day-ahead rates use the fleet's IFM awards, by hour.
• Real-time rates use the fleet's incremental real-time awards (RTPD award minus IFM award, floored at zero), by 15-minute interval, because real-time positions in the model are additional to day-ahead ones.

Acceptance rates limit volume, not price.

Expected throughput. Held capacity is converted into energy that moves SOC:

Spinning and Non-Spinning Reserve are rarely dispatched, so they carry small fixed values with no daily profile. The same values apply to the day-ahead and real-time legs of each service.

5.5 Data priority, fallbacks and minimum data

• Priority. CAISO-published prices are the only price source. Where CAISO republishes or corrects prices, the latest version available when the Benchmark is computed is used.
• Gaps in prices within a day. Where an interval is missing from an otherwise available price dataset, the most recent preceding value is used (or, at the start of a day, the following value).
• Gaps in fleet awards. Gaps of up to one hour (one day-ahead interval, or four real-time intervals) are linearly interpolated. Longer gaps are filled with the average award at the same time of day over the previous seven days. If no award data exists within seven days, no value is published for the affected days.
• Attenuation factors. Periods before the first published quarter (2023-Q4) use the first quarter's hourly profile. Periods after the latest published quarter use the latest quarter's profile until a new one is published.
• Minimum data. A Benchmark value is only determined for an operating day when all five price datasets in §5.1 are available for that day, and acceptance rates can be computed under the rules above. If any is absent, no value is published for that day until the data becomes available.

6. Modeling Methodology

6.1 What the model solves

The GDM answers one question: given the prices observed on a given day and the physical and regulatory limits of the asset, what is the highest revenue a well-run battery could have earned?

It allocates the battery's power and stored energy across energy (IFM, FMM and RTD) and ancillary capacity (Regulation Up, Regulation Down, Spinning and Non-Spinning Reserve, in the IFM and FMM). Revenue is the sum of these streams net of charging cost, subject to the constraints in §6.3–6.5.

6.2 Three-step optimization

The model solves CAISO's three market runs in sequence. Later steps hold the positions committed in earlier ones fixed.

1. Day-ahead step (60-minute granularity): IFM energy and the four ancillary products are optimized together.
2. FMM step (15-minute granularity): FMM energy and the four real-time ancillary products are optimized together within the remaining headroom.
3. RTD step (5-minute granularity): RTD energy is optimized within the remaining headroom.

Each step covers one operating day. The model has perfect foresight of energy and ancillary prices over the operating day.

6.3 Market stacking and physical limits

1. Power headroom: the sum of discharging ancillary positions (Regulation Up, Spinning, Non-Spinning) cannot exceed rated power, and nor can charging positions (Regulation Down).
2. Energy headroom: each discharging position must be backed by enough stored energy, and each charging position by enough empty capacity, to deliver at full power for the period CAISO's tariff requires:

   Service	Day-ahead (IFM)	Real-time (FMM)
   Regulation Up / Down	1 hour	0.5 hours
   Spinning Reserve	0.5 hours	0.5 hours
   Non-Spinning Reserve	0.5 hours	0.5 hours

3. Acceptance limits: capacity held in each service is capped at rated power times that interval's acceptance rate (§5.4).
4. Grid limits: combined charge and discharge cannot exceed the 100 MW import and export limits.
5. Daily cycling cap: total discharge throughput per day, including energy delivered through ancillary deployment, cannot exceed 1.0 cycles of usable capacity.

6.4 The battery's physical arrangement

• Round-trip efficiency (88%) is applied to charging: 1 MWh drawn stores 0.88 MWh.
• State of charge in each interval equals the previous level plus charging (net of efficiency) minus discharging, plus the expected energy flow from ancillary deployment (§6.5). The net flow within an interval is used, so offsetting trades within an interval incur no efficiency loss.
• SOC window is 0%–100%, starting and ending each operating day at 0%.
• Degradation is disabled.

6.5 Capacity vs deployed energy

The battery is paid the cleared capacity price (USD/MW/h) for each MW of ancillary capacity held. Holding capacity reduces the power and energy available to the energy markets. The model moves SOC by the held capacity multiplied by the expected throughput for that service and interval (§5.4). That energy counts toward cycling and, as in CAISO settlement, is settled at the RTD LMP: energy delivered for Regulation Up and reserves earns the RTD price, and energy absorbed for Regulation Down pays it. Regulation mileage is not paid.

6.6 Revenue calibration

A flat calibration factor of 90% is applied to simulated energy and ancillary revenue. The factor accounts for the gap between perfect-foresight modeled revenue and what a real battery earns, which has three sources:

• Foresight: the model sees prices a real trader does not have when offers are submitted.
• Availability: real assets have outages and maintenance; the representative asset is always available.
• Execution: trading frictions and slippage.

The value of 90% is supported by a comparison of GDM revenues with the realized revenues of the CAISO battery fleet, using fleet data from September 2025 to September 2026. It is applied after dispatch is optimized, so it does not change dispatch or the trade-offs between markets; it scales the result.

7. Expert judgment

7.1 Where judgment is exercised

The Benchmark is determined by the model from published data; there is no discretionary adjustment of individual daily values. Expert judgment is exercised in setting the methodology itself, specifically:

• the representative asset specification (§3.2);
• the construction of acceptance rates, including the treatment of real-time awards and the fallback rules for gaps (§5.4, §5.5);
• the Spinning and Non-Spinning expected-throughput values (§5.4); and
• the calibration factor (§6.6).

7.2 Keeping judgment consistent

These parameters and rules are fixed in the GDM's version-controlled codebase and applied identically to every operating day and every CAISO Benchmark. A parameter is changed only through a code change that is reviewed by a member of Modo Energy's modeling team other than its author and passes automated tests before release. Changes to parameters are treated as methodology changes under §11.

8. Market stress, disruption and missing data

Modo Energy's current practice is:

• Late or missing data. Where CAISO price data for an operating day is late or incomplete, Modo Energy waits for the data rather than estimating it. No value is published for that day until the minimum data in §5.5 is available. Short gaps in the fleet award data are filled as set out in §5.5.
• Extreme prices. Prices during scarcity, stress or other extreme events (for example, summer heat events) are used as published. The Benchmark is intended to reflect those conditions, so they are not removed or capped.
• Market suspension or rule changes. If CAISO suspends a market or changes its design, the Benchmark reflects the prices CAISO publishes for the affected period. Changes to the methodology needed in response are handled under §11.

9. Revisions to published values

Published values may be revised, for example where:

• Modo Energy's CAISO asset database is updated for a covered period;
• an error is identified in Modo Energy's processing of the input data; or
• a methodology change (§11) is applied to historical periods.

10. Methodology review

The methodology is reviewed internally by Modo Energy, including when CAISO changes its market design, when new data becomes available (for example, new attenuation factors or improved fleet data), or when changes are made to the GDM. Because CAISO publishes fleet-level data, reviews may include comparing the Benchmark with the ME BESS CAISO System Benchmark. Any changes that result from a review are made under §11.

11. Methodology changes

Changes are classified as material or non-material. A change is material where it may meaningfully affect benchmark values, the inputs used, or what the benchmark measures. Material changes are announced with not less than 30 days' written notice before they take effect, together with an impact analysis.

Non-material changes are corrections and clarifications that do not alter the substance of the methodology. They may be made without prior consultation and are disclosed as soon as practicable.

Emergency changes needed to protect the integrity of the benchmark or to comply with law may be made without prior consultation, with notice given as promptly as circumstances permit.

All changes are recorded, with their effective date, in the change log (Appendix I).

12. Limitations

• Simulated, not observed. The Benchmark is not derived from the transactions or revenues of any real asset. It may differ materially from the ME BESS CAISO System Benchmark and from any individual asset.
• Foresight and calibration. The model optimizes with perfect foresight. The 90% calibration factor is a single flat value fitted over one year of fleet data; actual foresight losses vary over time.
• Thin and concentrated ancillary markets. CAISO's ancillary requirements are small relative to its battery fleet, and batteries win nearly all awards. Acceptance rates are fleet-average, pro-rata values: an individual asset may clear more or less than its share. The asset is assumed not to affect clearing prices.
• Single AS region. All benchmarks use AS_CAISO_EXP ancillary prices, so locational differences in ancillary prices are not captured.
• Ancillary activation. Regulation throughput uses CAISO's quarterly hourly attenuation factors, which are averages and do not reflect specific days. Reserve throughput uses small fixed values. Periods before 2023-Q4 use the 2023-Q4 profile.
• Asset assumptions. The asset is always available, does not degrade, and starts and ends each day empty. Real assets differ on each point.
• Excluded revenues. The revenues and costs listed in §4.2, including resource adequacy, are not captured. RA is a large part of CAISO battery revenue in practice.

Appendix I — Change log

Disclaimer

This document, including the methodologies and benchmarks described herein, is the proprietary work of MODO ENERGY LIMITED ("Modo Energy") and is provided solely for informational purposes. These benchmarks are designed for use in financial analysis, benchmarking, and decision-making. However, they do not constitute investment advice or a recommendation regarding any specific financial instrument, asset, or strategy.

While Modo Energy strives to ensure the accuracy, reliability, and transparency of the benchmarks and methodologies, all information is provided "as is", without any express or implied warranties, including but not limited to warranties of merchantability or fitness for a particular purpose. Users should be aware that the benchmarks are derived from publicly available market data that may be subject to revisions, delays, or inaccuracies, and that the CAISO virtual asset Benchmarks are simulated from a representative asset rather than observed from underlying transactions. Past performance is not indicative of future results, and external factors such as regulatory changes, market conditions, and asset-specific characteristics may impact benchmark performance.

Modo Energy encourages users to conduct their own due diligence and consult with qualified financial professionals before making any investment or operational decisions based on the benchmarks or methodologies herein. Modo Energy disclaims any liability for direct, indirect, incidental, or consequential losses or damages arising from the use of the benchmarks, methodologies, or related data.

It is not possible to invest directly in a benchmark. Benchmarks are intended to represent performance references, and exposure to an asset class represented by a benchmark may be available only through separate investable instruments. Modo Energy does not sponsor, endorse, or manage any financial products that aim to track the performance of its benchmarks.

All intellectual property rights to the benchmarks and methodologies are owned by Modo Energy. Unauthorized use, reproduction, or redistribution of this document, in whole or in part, is strictly prohibited without prior written consent from Modo Energy.

This document and the benchmarks it describes are subject to updates and revisions. For further information, including licensing inquiries, please contact Modo Energy directly.

By accessing or using this document, you acknowledge and accept the terms of this disclaimer.