BESS Dispatch Optimization for IPPs: How SoH Errors Create Revenue Leakage
Singapore's ancillary services market, specifically Primary Reserve, Contingency Reserve, and Regulation services under EMA's USEP framework, has been the main revenue driver for the wave of BESS deployments commissioned between 2020 and 2024.
On paper, the revenue model is simple: commit capacity, deliver response within specification, collect payment. In practice, profitability is governed by how accurately you know your asset's true State-of-Health, and many operators do not know it as well as they think.
The BMS SoH to revenue pipeline
Your dispatch software relies on BMS-reported SoH to calculate the usable energy window for each trading interval. If your 20 MWh BESS reports 85% SoH, your dispatch engine treats it as having 17 MWh usable capacity. That determines:
- Regulation bid volume: how many MW of regulation up/down you commit to EMA
- Dispatch depth: how aggressively you cycle in energy arbitrage windows
- State-of-charge management: where you park the battery between dispatch events
When BMS SoH is wrong, and for assets older than two years it often is, every downstream calculation in your dispatch stack inherits the error.
Scenario A: BMS Overestimates SoH
This is the more dangerous case. BMS reports 85% SoH, actual capacity is 74%.
Regulation bids: You commit 8 MW regulation capacity. When a regulation-up event requires sustained discharge, your asset reaches its lower SoC limit before the event window closes. You fail to deliver, which can cost you the interval and trigger a non-compliance flag.
Energy arbitrage: You schedule a full charge from 0900 to 1200 during the low USEP price window and plan to discharge from 1400 to 1700 during the peak. But actual capacity is lower, so the projected discharge falls short.
Degradation acceleration: If the system believes usable capacity is 17 MWh but actual is 14.8 MWh, it charges at an effectively higher C-rate to fill the assumed window. That directly accelerates fade and plating risk.
Scenario B: BMS Underestimates SoH
Less dangerous, but expensive. BMS reports 72% SoH, actual capacity is 81%.
Stranded capacity: You are holding usable energy in reserve that the dispatch engine never touches. For a 20 MWh BESS at an arbitrage delta of S$80/MWh, that can mean about S$47k a year in unrealized revenue.
Conservative cycling: If your SoC management is set to protect a degraded asset, you operate at shallower depth of discharge than necessary and reduce yield further.
Both scenarios show up often once a battery has been cycling for a couple of years. The size of the gap depends on chemistry, duty cycle, temperature, and how the BMS was calibrated, which is exactly why operators should validate SoH from telemetry instead of assuming the dashboard number is close enough.
Frequency Regulation: Where SoH Errors Are Most Expensive
Singapore's ancillary services require fast response. Primary Reserve requires injection within 30 seconds of a frequency event. For BESS, that is mechanically achievable. The harder part is capacity reservation: EMA requires that reserved capacity be continuously available.
This creates the reservation-versus-dispatch tension:
If your BMS overstates SoH:
- You reserve more capacity than physically exists
- A severe frequency event requiring maximum injection cannot be fully delivered
- Non-delivery under ancillary service contract terms triggers financial penalties
If your SoH is correct:
- You reserve precisely the capacity you can deliver
- You release the remaining SoC window for energy arbitrage
- Dispatch software runs a tighter, more profitable intraday stack
The Singapore electricity market has deepened. USEP spreads have widened, and the number of BESS bidders has not grown proportionally with overall renewables capacity. Operators with accurate SoH have a structural advantage in regulation bidding because they can commit closer to actual capacity and still deliver reliably.
The Degradation-Aware Dispatch Framework
Correcting for SoH error in your dispatch stack requires three components:
1. Physics-Based SoH Baseline (Oxaide Verify)
A one-time forensic audit of your full cycling history produces a calibrated SoH baseline and degradation mode classification. This replaces BMS SoH as the initial operating point in your dispatch model.
Typical engagement: 5 business days, defined review scope.
2. SoH Trajectory Model (Oxaide Horizon)
Once you have a calibrated baseline, real-time SoH tracking through Oxaide Horizon's streaming analytics engine maintains accuracy across future cycling. The model updates dynamically as each cycle completes:
- Adjusts usable SoC window as capacity changes
- Detects when cycling protocol is accelerating degradation vs. baseline trajectory
- Triggers operational alerts when lithium plating risk score increases
3. Dispatch Integration
Horizon exposes a lightweight API that your Energy Management System or trading platform can query for:
- Current True SoH (float, 0-1)
- Usable energy window (MWh)
- Recommended SoC ceiling/floor for protective operation
- Thermal risk flag (boolean + risk score)
For most commercial dispatch software stacks, integration is a configuration change, not a development project.
Revenue Impact Quantification
Based on typical 10 MW / 20 MWh BESS assets operating in Singapore's ancillary services market:
| Error Type | SoH Gap | Annual Revenue Impact |
|---|---|---|
| BMS overestimate (non-delivery penalties) | +10 percentage pts | S$85k to S$160k losses |
| BMS overestimate (accelerated degradation cycle cost) | +10 percentage pts | S$60k to S$120k degradation cost |
| BMS underestimate (stranded capacity) | -8 percentage pts | S$35k to S$55k unrealized revenue |
Correcting a +10 point SoH overestimate through physics-informed dispatch removes S$145k to S$280k in annual risk exposure for a standard 10 MW asset. A proper baseline plus a disciplined monitoring layer is materially smaller than any of these figures.
Singapore Regulatory Context
EMA's Market Rules require BESS operators to maintain accurate metering and capacity declarations. Incorrect capacity bids resulting from BMS SoH errors are not a regulatory defence. The asset owner is responsible for the accuracy of those declarations.
As EMA increases BESS penetration targets and ancillary services procurement, with the Forward SG grid plan projecting significantly higher BESS capacity by 2030, enforcement of declaration accuracy is likely to tighten. Operators relying on uncalibrated BMS SoH figures face both financial and compliance risk.
The BESS operators who outperform in Singapore's electricity market over the next five years will be the ones who manage assets based on physics, not BMS approximations. State-of-Health accuracy is not a back-office data-quality question. It sits directly in your revenue P&L.
Related service pages:
Commission a SoH Calibration Audit → | Explore Horizon Real-Time Monitoring