---
id: VOLT-HOME-WP-032
title: "How degradation cost changes the optimal home-battery schedule"
slug: how-degradation-cost-changes-the-optimal-home-battery-schedule
description: "A wholesale-only BESS-v2 break-even estimate shows how little normalized daily value is available to absorb cycling wear before schedule economics disappear."
published: 2026-08-30
cluster: "Home batteries"
status: measured
evidence_url: /research-data/home-papers/how-degradation-cost-changes-the-optimal-home-battery-schedule.json
figure_url: /research-media/home-papers/how-degradation-cost-changes-the-optimal-home-battery-schedule.webp
figure_alt: "Chart for How degradation cost changes the optimal home-battery schedule: median daily degradation-cost break-even, shown as P10, Median, P90."
source_ids:
  - energy-informatics-storage
  - acer-retail-2025
  - iea-electricity-2026
  - dynamic-tariff-viability
  - volt-research-content
peer_reviewed: false
---
## Abstract

Battery dispatch can look attractive before wear is priced and unattractive after it is. The frozen evidence for this paper reports a median daily degradation-cost break-even of 0.067935 EUR per kW-day across 16,368 normalized BESS-v2 observations. The evidence interpretation is precise: degradation above this normalized amount removes the index’s gross wholesale value. That is a break-even accounting result, not a measured change in the interval-by-interval schedule and not a device-life estimate.

BESS-v2 evaluates a normalized day-ahead battery under perfect foresight. It uses wholesale energy only and excludes retail taxes and network charges. The result therefore answers a narrow question: how much daily degradation charge can the indexed gross value absorb under the frozen convention? It does not answer what a battery’s wear actually costs, because the evidence contains no chemistry, temperature, depth-of-discharge, state-of-health, warranty, or replacement-cost model. Nor does it show the schedule selected after adding a degradation penalty. The appropriate conclusion is that even modest wear assumptions can be decision-critical and must be included before interpreting a perfect-foresight spread as economically available.

## Plain-language answer

The central recorded break-even is €0.067935 per kW of normalized power per day. If the degradation charge assigned to the indexed operation is higher than that amount, the median gross wholesale value is exhausted under this study’s convention. If it is lower, some indexed value remains before other omitted costs. The number is daily and normalized; it is not an annual payback, a household bill saving, or measured site cashflow.

Degradation changes a schedule by attaching a cost to throughput. Without that cost, an optimizer can take every spread that is positive after efficiency loss. With it, small spreads should be skipped because the extra cycle consumes scarce battery life. The evidence does not expose the resulting dispatch path, so this paper cannot say which hours were dropped or how cycle count changed. It can say that the median daily allowance is small enough that wear cannot be treated as an afterthought. Taxes, network charges, contract spreads, standby use, and forecast error would further reduce the amount available to pay for degradation.

## Research question

The registered question is how degradation cost changes the optimal home-battery schedule. The measurable evidence narrows that to a break-even question: what normalized daily degradation amount equals the gross wholesale value of the BESS-v2 index? This distinction matters. A complete schedule comparison would require at least two optimized paths—one without degradation and one with a specified marginal wear curve—plus interval-level differences. The published evidence supplies a scalar break-even, not those paths.

The economic mechanism is nevertheless clear. A throughput penalty raises the minimum acceptable price spread. The optimizer should reserve energy capacity and cycle budget for higher-value intervals, potentially reduce partial cycling, and avoid a second marginal cycle. Whether that produces fewer starts, shallower depth of discharge, or different hours depends on asset and market details not in the evidence. This paper therefore treats the break-even as a screening constraint on optimization, not as proof of a particular control pattern.

## Data and provenance

The analysis uses the common frozen corpus snapshot, SHA-256 `7e97489fc8528c8cc8c38830e05b48d949ce1f67b98575dff26f5d7c321e4c67`, under a SELECT-only transaction and 180-second statement timeout. The registered contracts are `day_ahead_prices`, `bess_index_daily`, `bess_forecast_daily`, `generation_mix`, and `capture_stats`. Daily prices span 2021-01-01 to 2026-08-29, detailed intervals span 2025-10-01 to 2026-08-29, and the long-history boundary spans 2015-01-01 to 2026-08-29. The publication cut-off is 2026-08-30T00:00:00Z.

The primary sample size is 16,368, the metric is median daily degradation-cost break-even, and the unit is EUR per kW-day. The code hash is `57c57de79cdab2b5b6d6c54c485cb5162598c5ba0b0bfe995da40d75e6c52ba9`; protocol hash, `adb36bf6b447af9f96339249b8becaefc20422499cca1977242866347a97bd4b`; registry hash, `7bcb91d7476d0a69fe9fa75a5c7782f8117e0153f82f9112b7e1d307d3943717`; source-registry hash, `07949550ac443ff673fda5c0209b99f137544f3ffecf6775f109bb9d09663bd6`. These identities bind the public aggregate to the frozen analysis but do not substitute for a device-specific wear model.

## Method

BESS-v2 normalizes the indexed asset per unit of power so market opportunities can be compared without selecting a named installation. The reported EUR per kW-day unit expresses how much gross indexed value is available for a degradation charge at that scale. Multiplying or dividing by a device rating is only unit scaling. It does not account for usable capacity, temperature control, inverter efficiency, warranty terms, or how degradation changes nonlinearly with depth and rate.

The index uses perfect foresight of the realized day-ahead curve. That means the gross value is generated by interval choices made with information unavailable at the real scheduling clock. It is therefore an upper-bound-like input to the break-even calculation. The method isolates the normalized daily amount that would reduce gross wholesale value to zero. Retail taxes and network charges are excluded. Within-family Holm correction governs inferential claims, but this descriptive result makes no unadjusted significance claim. No causal interpretation is attempted.

## Results

The primary evidence value is 0.067935 EUR per kW-day. The exact precision is retained here because it is the frozen machine output, although a practical model would not know degradation cost to six decimal places. The evidence states that degradation above this normalized amount removes the index’s gross wholesale value. The result should be interpreted as the median allowance across 16,368 index rows, not a minimum, maximum, warranty limit, or annual average for a specific home.

The regenerated evidence reports `bootstrap_95_interval: null` and no interval method for this estimand. The figure publishes P10 = 0.013634799999999999, median = 0.067935, and P90 = 0.28450960000000014 EUR per kW-day. These are distribution quantiles, not uncertainty bounds around the median. The defensible empirical statement is the recorded point estimate, sample size, unit, assumptions, interpretation, and figure distribution; no confidence interval is claimed.

The finding does not directly reveal schedule changes. It implies that a degradation-aware schedule should reject marginal opportunities whose gross contribution is below the applicable wear charge, but the JSON contains no cycle-level counterfactual. Any count of skipped intervals, cycles, or depth changes would be invented and is intentionally absent.

## Robustness and placebo checks

A robust degradation study would test several wear representations: a constant throughput fee, a depth-dependent curve, separate calendar and cycling aging, and stress by temperature and C-rate. It would compare schedule paths and gross-to-net value on identical price rows. The current public evidence reports none of those arms, so this paper does not claim they were run. The within-family Holm policy protects inferential families, but it cannot repair a missing device model or supply uncertainty for an estimand whose interval is not reported.

Useful placebo checks would include setting degradation to zero and recovering the original BESS-v2 schedule, setting it above the break-even and verifying that optional marginal dispatch disappears, and shuffling degradation charges across otherwise identical assets to confirm that only the declared cost changes the decision. Forecast-based and perfect-foresight schedules should also be separated. These are requirements for future validation, not hidden results. The present robustness comes from frozen provenance, explicit accounting boundaries, and refusal to invent an interval where none is reported.

## Limitations

Actual degradation is path-dependent. Lithium-ion wear depends on chemistry, state of charge, depth of discharge, charge and discharge rates, temperature, dwell time, age, and control strategy. A constant daily break-even does not model those interactions. It also does not price warranty thresholds, replacement timing, salvage value, or financing. Two batteries with the same power rating can have different usable energy and wear per unit of throughput.

Perfect foresight inflates schedule quality relative to a controller choosing from a forecast. Wholesale-only accounting omits taxes, network charges, supplier mark-ups, import-export asymmetry, and fixed fees. The normalized index is not a household bill model. The evidence does not report schedule paths, cycle counts, or an uncertainty interval for the estimand. Consequently, the paper cannot claim that 0.067935 EUR per kW-day is a universal degradation cost, only that it is the frozen median gross-value break-even under BESS-v2.

## Practical implication

Any battery optimizer should make degradation explicit rather than burying it in a post-processing adjustment. The decision rule should compare each forecasted opportunity with efficiency loss, a documented marginal wear cost, tariff and network effects, and uncertainty. If the controller cannot justify its wear assumption, it should expose a sensitivity range and avoid marginal cycling. A high gross spread can survive a conservative wear charge; a small spread may not.

For evaluation, use the reported 0.067935 EUR per kW-day as a benchmark scenario, not as a recommended parameter. Re-run the schedule with device-specific cost curves and the information available at the decision clock. Report both the gross perfect-foresight ceiling and attainable forecast-based result. This prevents an attractive normalized index from being mistaken for an observed site outcome and shows exactly which assumptions make the schedule change.

## Reproducibility

The public evidence is `/research-data/home-papers/how-degradation-cost-changes-the-optimal-home-battery-schedule.json`. Check that the status is `measured`, the point estimate is 0.067935 EUR per kW-day, sample size is 16,368, no interval is reported, and the interpretation and assumptions match this text. The manifest binds the JSON to SHA-256 `580348cbc5aefb371355d18a08bc4808ddb006ab997566bf2dfc6003837ae79f` and the figure to `05ebace87f080057b3f0b4e4cb2f20a76b48ab1bd0ac9ddea3361331e970f879`.

Reproduction must use the frozen snapshot, protocol, registry, code hash, time windows, and publication cut-off stated above. It should preserve the EUR per kW-day unit and avoid silently converting the break-even into a cycle-life estimate. Any new schedule analysis must disclose its degradation curve and compare paths on identical rows. Public use remains subject to `/legal/data-licensing` and [Voltcast data licensing and redistribution](https://github.com/ossedk/voltcast/blob/main/docs/voltcast/LICENSING.md).

## Disclosure

Analysis and drafting were model-assisted; sources, code identities, assumptions, evidence hashes, and unresolved inconsistencies are disclosed. This public working paper is not peer reviewed. It is not trading advice, financial advice, an investment recommendation, or a warranty analysis. Voltcast has no live traders or live capital. No schedule, market action, or purchase is authorized by this descriptive result.

## References

- `energy-informatics-storage` — Energy Informatics. [Risk and reward: evaluating household energy storage for optimizing demand-side flexibility under dynamic tariffs](https://doi.org/10.1186/s42162-025-00602-9). Kind: peer-reviewed.
- `acer-retail-2025` — ACER and CEER. [Rewarding flexibility: How retail contract choice can help unlock consumer flexibility](https://www.ceer.eu/wp-content/uploads/2025/11/ACER-CEER-2025-Retail-monitoring.pdf). Kind: official.
- `iea-electricity-2026` — International Energy Agency. [Electricity 2026](https://www.iea.org/reports/electricity-2026). Kind: official.
- `dynamic-tariff-viability` — Advances in Applied Energy. [Assessing the conditions for economic viability of dynamic electricity retail tariffs for households](https://doi.org/10.1016/j.adapen.2024.100174). Kind: peer-reviewed.
- `volt-research-content` — Voltcast. [Voltcast Research Content Plan](https://github.com/ossedk/voltcast/blob/main/docs/voltcast/RESEARCH-CONTENT-PLAN.md). Kind: canonical.