Frequently asked questions

Market coverage is driven by customer requests and regulatory activity. HERO supports multiple markets including Spain and North American ISOs. Additional European markets will be available powered by DNV’s Power Price Forecasts published in DNV’s Power Analytics platform.

European markets (using DNV's Power Price Forecasts):

• Spain (revenue stacking available since 2025 for the day-ahead and balancing markets)
• Portugal (Planned for 2026)
• Italy (Planned for 2026)
• France (Planned for 2026)

North American markets:

• CAISO (available)
• ERCOT (available
• NYISO (available)
• IESO (available)
• ISONE (available)
• PJM (not available)
• MISO (not available)
• SPP (not available)
• AESO (not available)

We continue expanding toward additional markets. If a missing market is limiting your evaluation of our product, please contact us.

The value streams and the time resolution at which they are currently modeled in HERO are:

• Day-ahead energy markets (hourly and 15-minute)
• Balancing markets (hourly and 15-minute):
  • aFRR (Automatic Frequency Restoration Reserve) power and energy
  • mFRR (Manual Frequency Restoration Reserve) energy
• Ancillary services (for North American markets):
  • Regulation up/down (hourly)
  • Spinning reserve (hourly)
  • Capacity (Annual price $/kW-month and EUR/MW)

We constantly strive to improve the system topologies that are offered through HERO. A list of system topologies that we intend to offer are:

• Stand-alone storage (available)
• DC-coupled PV (available)
• AC-coupled with PV on shared bus (available)
• AC-coupled with wind on shared bus (available)
• DC-coupled with PV and AC-coupled with wind (under development)
• AC-coupled with PV and wind (under development)

Yes, we do consider POI limitation i.e., the output of a hybrid plant that is modeled in HERO will not exceed the interconnection limit.

Currently, HERO supports one-hour resolution data only for renewable production data.

Yes, the user can upload a file to specify the percentage of charging from renewables.

HERO uses MILP (mixed-integer linear programming) approach in the optimization algorithm.

HERO generates a dispatch by assuming a perfect foresight. However, prior to calculation of revenues, we apply income adjustments to account for uncertainties. While this approach is not perfect, this is the closest accurate process that we have come up with.

We are also contemplating offering a company level profile where the administrator can set income adjustments of their choice by ISO.

Our team also has been developing an uncertainty analysis approach which is a post-processing mechanism. While this is not currently integrated into HERO, it is on our roadmap. We do offer this service as a part of our advisory engagements. If you wish to learn more, please reach out to us or view this webinar - Financing the next generation of merchant energy storage projects

HERO has undergone a UI/UX refresh in Q3/2023. As a part of this feature, we have introduced our new “Constraint Designer” feature. This allows a user to place constraints down to the hourly resolution.

Currently, system availability is hardcoded with a standard assumption (97% system availability @POI). We will be offering the ability for these inputs to be user provided soon.

While HERO is capable of handling finer resolution of real-time data, we currently do not offer this feature. We believe that using finer resolution of real-time price forecasts will increase the uncertainty and increase the time of optimization without proportional pay-off.

Similarly, moving to hourly granularity would essentially be taking an average of the 15-minute price. This could reduce an intra-hourly opportunity for the BESS to take advantage of a 15-minute price spike vs participation in other products.

Essentially you would lose some precision in the way the HERO algorithm is making decisions. Depending on the price forward, namely how volatile the 15-minute prices vs hourly prices are would dictate the impact on the switch.

According to DNV, using 15-minutes for real-time data is a reasonable tradeoff and we will continue to assess this as we monitor and study the market further.

The charge and discharge efficiency are calculated as a portion of BESS losses. Typically, this is just seen as one part of the RTE with others being DC Line efficiency, DC Converter efficiency, and AC inverter efficiency. All of these are requested in HERO to calculate the full RTE which is a product of all system efficiencies. These are used in optimization when accounting for energy sufficiency to satisfy schedules, charge from grid, SoC, etc.

The shadow price is a lever we use to control the annual cycle. It is used in the optimization formulation only and doesn’t result in any actual cost. Instead, it works like a hurdle rate in modeling, i.e., only cycle the battery if it can make more than $/kWh in return. Because it is not an actual cost, if there are any O&M costs associated with operating the battery, these should be separately included in the OpEx in the financial model. Currently shadow price is applied on discharge throughput only. But again, because it is an arbitrary cost without any physical meaning, it would work the same if it’s applied to both charge and discharge (with the cost being approximately half if it’s applied on both).

In terms of configuration, shadow price is largely dependent on the price profiles, hence there is not a single good value to use. Currently to control the annual cycle within the cycling cap, an iterative approach is used. We typically recommend the following steps:

• Run the scenario with shadow price = 0
• If the resulting # of cycles is within the annual cap, then there is no need to increase the VOM to make the battery cycle less, and the results will be final. VOM cannot be negative.
• If the resulting # of cycles is more than the annual cap, then the shadow price must be increased to bring the cycles down. Generally, we see a somewhat linear relationship between the # of cycles and the shadow price, and hence, you could try step 2 to 4 below.
• Run the scenario with shadow price = 10

The estimated shadow price (vom_t) to reach the target number of cycles (cycle_t) would be:

vom_t=(10-0)/(cycle_10-cycle_0 )*(cycle_t-cycle_0) 
where:

• cycle_10 – represents the # of cycles at vom=10
• cycle_0 – represents the # of cycles at vom=0

Run the scenario with shadow price = vom_t. Some fine tuning might be needed if the resulting # of cycles are still far away from the cycle target. For fine tuning, you could either increase/decrease the shadow price in small increments, or do a linear regression based on all the points/trials you have.

Also, note that we now offer an alternative mechanism to control the cycling of the battery where the user can define the annual cycles, they want to limit the BESS to each year. 

Yes. HERO leverages DNV’s Power Price Forecasts for selected countries, delivered through DNV’s Power Analytics platform. These forecasts are produced by DNV’s energy market experts and include long‑term power price outlooks, balancing market signals (including aFRR and mFRR where available), and market‑specific assumptions. DNV currently provides power price forecasts for a growing set of international markets, including several European countries where HERO revenue stacking is supported or planned. These forecasts are used directly within HERO for market simulations and revenue estimation. For more information or to request access to Power Price Forecasts for additional geographies, please contact us.