Pith. sign in

REVIEW 5 major objections 6 minor 3 references

Emission-Aware Operation of Electrical Energy Storage Systems

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A regression-based marginal emission intensity signal lets storage dispatch shift to low-emission hours, earn carbon credits, and out-earn price-only operation in an Ontario case study.

desk verdict A cleanly written carbon-aware storage dispatch paper whose central MEI estimates are unvalidated correlations, including a physically implausible negative emission intensity in the lowest demand segment. read the letter →

arxiv 2506.16454 v1 pith:YECW5H57 submitted 2025-06-19 eess.SY cs.SY

Haotian Yao , Vahid Hakimian , Mostafa Farrokhabadi , Hamidreza Zareipour This is my paper · ORCID
classification eess.SYcs.SY
keywords energystoragesystemsmarginalemissionintensitycarbonaccountingperformancecreditscarbon-awaredispatchresidualdemandmarketparticipation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that a storage operator can learn the grid's real-time marginal emission intensity (MEI) by regressing the output of load-following generators — gas, hydro, and net imports — against residual demand, and that this signal lets batteries of any size be dispatched to avoid emissions and earn tradable emission performance credits (EPCs). If correct, storage no longer needs arbitrary weighting coefficients to balance profit and carbon; the carbon price enters the dispatch objective directly through a measured, hour-by-hour emission intensity. The authors support this with a six-month Ontario case study in which a combined price-and-carbon dispatch earns the highest operational revenue of the three tested strategies while also reducing emissions. The point of the framework is to make storage a verifiable participant in compliance carbon markets rather than an invisible load.

What carries the argument

The load-bearing object is the segment-wise marginal supply share λ_{n,s}: the slope of a piecewise-linear approximation of the cubic regression that maps each hour's residual demand to the output of gas, hydro, and net imports. Multiplied by each resource's emission factor and summed, these slopes give the marginal emission intensity m_s for each of the 15 residual-demand segments. This is the quantity that converts the carbon market price into a per-MWh operating signal for storage, so the optimization in (5) becomes a pure arbitrage over the combined price λ^G_t + λ^C_t ρ^G_t.

What would settle it

Compute the actual marginal unit from Ontario dispatch data for a sample of hours (e.g., the gas plant or import schedule at the margin) and compare the realized change in emissions per MWh with the MEI m_s predicted by Table 1; if the realized marginal emission rate differs systematically, or if the per-segment slopes in Table 1 do not reproduce out-of-sample residual-demand responses, the claimed emission reductions would not materialize.

Watch

Extended reading notes

Core claim

The central discovery is a practical method for real-time MEI that respects how grid operators actually dispatch marginal units. Hourly residual demand is defined as total Ontario demand minus nuclear, wind, solar, and bioenergy output; cubic regressions of gas generation, hydro generation, and net imports against residual demand are fitted on six months of IESO data, then approximated piecewise-linearly in 15 segments. The slope of each resource in each segment, λ_{n,s}, is its marginal supply share, and the segment's MEI is m_s = Σ λ_{n,s} e_n, where gas emits 0.37 tCO2e/MWh, imports 0.44 tCO2e/MWh, and hydro zero. The sum of the three slopes stays near one across segments, which the paper reads as confirmation that the marginal demand is fully assigned. The MEI feeds a rolling-horizon ESS dispatch that maximizes revenue from electricity prices plus carbon price times MEI, so storage charges when the grid is clean and cheap and discharges when it is dirty and expensive. In the Ontario case study the dual-signal strategy (Case 3) earns $17.67k versus $17.30k for price-only dispatch while cutting 1.94 tCO2, where the price-only strategy actually increases emissions by 6.62 tCO2.

Load-bearing premise

The regression slopes λ_{n,s} estimated from six months of Ontario data are assumed to be causal marginal dispatch responses that remain valid for carbon accounting over the simulated period.

Editorial extensions

If this is right

  • Storage operating on the combined signal can out-earn price-only arbitrage while reducing emissions; in the paper's Ontario simulation, Case 3 revenue exceeds Case 1 by about $370 over six months and avoids 1.94 tCO2 where Case 1 adds 6.62 tCO2.
  • A carbon price alone triggers carbon-aware dispatch regardless of its magnitude; the paper's sensitivity analysis shows the emission surface for Case 2 stays flat along the carbon-price axis, so only the presence of the signal matters when electricity price is ignored.
  • Emission reductions scale most strongly when carbon price and storage capacity rise together; the diagonal of the sensitivity surface drops faster than either axis alone, implying coordinated policy and infrastructure investment is needed rather than either lever alone.
  • The MEI signal is jurisdiction-specific but the method transfers: any system operator can fit the same residual-demand regressions from public dispatch data and produce segment MEIs for its own fuel mix.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because the MEI is built from historical dispatch slopes, a jurisdiction with fast-changing fuel prices or new gas capacity would need periodic refits of the 15 segment slopes; the paper's six-month data window leaves open how quickly the λ_{n,s} values drift.
  • The method implicitly credits storage for emissions avoided at the margin, but not for the emissions embedded in battery manufacturing; a full life-cycle accounting would change the break-even carbon price and the EPC quantity attributed to each cycle.
  • A natural test is to compare this regression MEI against a unit-commitment-based MEI (e.g., from a production-cost model) over the same hours; agreement would validate the slopes as causal, disagreement would locate where the piecewise approximation loses the true marginal unit.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 6 minor

Summary. The paper proposes an 'emission performance credits' (EPCs) framework that lets energy storage systems (ESS) participate in carbon markets by trading the emissions they avoid through time-shifting. The central technical contribution is a real-time marginal emission intensity (MEI) for the grid, estimated by regressing the outputs of gas, hydro, and net imports on residual demand using six months of Ontario data (October 2024–April 2025), then approximating the fitted curves with 15 piecewise-linear segments. The MEI is used in a linear optimization (Eqs. 5–7) that co-optimizes electricity arbitrage and emission-credit revenue. Three cases are simulated: price-only, carbon-only, and dual-signal dispatch, for 1 MWh and 4 MWh ESS. The paper reports that dual-signal dispatch achieves the highest revenue (17.67×10^3$) with modest emission reduction (1.94 tCO2), while carbon-only dispatch gives the largest reduction (9.75 tCO2) but low revenue. A sensitivity analysis shows emissions fall most when ESS capacity and carbon price increase together. The authors claim the framework is practical, scalable, and enables ESS to generate tradable credits under compliance carbon markets.

Significance. If the proposed MEI were valid as a real-time marginal emission signal, the framework would be a useful and novel step toward incorporating ESS into carbon markets, and the optimization model (Eqs. 5–7) is coherent and computationally tractable. The paper also makes a laudable effort to base the analysis on publicly available system data and to report numerical results for several cases. However, the significance is conditional on the validity of the MEI, which is not established: the estimated λ_n,s are correlations rather than causal dispatch responses, the regression fits have low to moderate R², there is no out-of-sample or uncertainty analysis, and the reported emission reductions are computed from the very same MEI that drives the dispatch. These gaps are load-bearing because every claimed reduction and every EPC value derives from the fitted MEI.

major comments (5)
  1. [§2.1, Table 1] The slopes λ_n,s in Table 1 are estimated by ordinary least-squares regressions of gas, hydro, and import output on residual demand, but this does not identify causal marginal dispatch responses. Residual demand is jointly determined with prices, generator availability, and external market conditions, so the OLS coefficients are confounded. The near-unity sum of slopes across segments, cited as verification, is close to an accounting identity because gas plus hydro plus imports approximates residual demand by construction. More directly, Table 1 reports negative gas slopes in segments 1 and 15 and a negative MEI of -0.053 tCO2/MWh in segment 1; a negative marginal emission intensity implies that increasing demand reduces emissions, which is physically implausible for a system with fossil marginal generation and signals regression artifacts rather than dispatch behavior. The paper should provide a causal identification strategy, compare against unit-level dispatch data, or validate against a production cost model before the MEI can be treated as a real-time emission signal.
  2. [§2.1, footnote 2 and §3] Footnote 2 promises that 'the implications of using total hydro generation, rather than isolating peaking hydro, on the residual demand relationship are examined in a subsequent section,' but no such examination appears in Section 3 or Section 4. The hydro regression (R²=0.437) mixes baseload hydro with the dispatchable component, and the single fitted curve is then used to assign a marginal supply share λ_hydro,s in every segment. Conflating baseload and marginal hydro biases the MEI values in Table 1. The authors must either provide the promised analysis or delete the claim; without it, the hydro contribution to the MEI is unsupported.
  3. [§2.2, Table 2] The emission reductions reported in Table 2 are computed using the same ρ^G_t that appears in the objective function (Eq. 5): the dispatch is optimized to minimize the cumulative sum of ρ^G_t times net grid draw, and the 'emission reduction' is then presumably the change in that same sum relative to a baseline. This means the headline result is determined by construction from the fitted MEI and cannot serve as independent evidence that emissions actually fell. To substantiate the claim of real emission reduction, the paper needs an external emissions accounting: for example, recomputing emissions using unit-level generation and measured stack emissions, or at minimum using an independently established average emission intensity, and showing that the dispatch changes those quantities. Without such validation, the EPC quantities and revenue figures in Table 2 are not credible.
  4. [§2.1, Eq. (2)] The import emission factor is fixed at 0.44 tCO2/MWh and applied only in segments 14 and 15, with no sensitivity analysis or justification for that value over the study period. The MEI in those high-demand segments is dominated by the import share (λ_import,14=0.671, λ_import,15=0.880), so the claimed MEI of 0.361 and 0.369 tCO2/MWh would change substantially if the import factor differs from 0.44. Moreover, no confidence intervals are reported for any λ_n,s or m_s, even though the gas and import regressions have R² of 0.824 and 0.314, respectively. The paper should include uncertainty quantification and a sensitivity analysis on the import emission factor.
  5. [§1, footnote 1 of §2] The entire economic incentive of the proposed framework rests on the assumption that EPCs can be sold to regulated entities under compliance carbon markets. No existing regulatory program is cited that recognizes emission-shifting by ESS as a source of tradable compliance credits; the cited regulations (European Union, 2023; Government of Alberta, 2023) apply to direct emitters and do not mention ESS or EPCs. The authors should either identify an actual compliance mechanism that accepts such credits, or explicitly frame EPCs as a proposed instrument rather than an available market product. As written, the 'participation in carbon markets' claim is an unsupported premise.
minor comments (6)
  1. [§2.1] Typo: 'Accroding' should be 'According' in the first sentence of Section 2.1.
  2. [§2.1, footnote 3] Typo: 'pubic' should be 'public' in the IESO data directory footnote.
  3. [§3.2] Figures 5a–5f are cited as 'Figures 5' in the text, but the individual panels are not referenced consistently; please refer to specific panels (e.g., 'Figure 5d') when discussing 6–8 AM behavior.
  4. [§3.1] The paper assumes perfect foresight for electricity prices and MEI over the optimization horizon ('access to accurate predictions'), which is a strong assumption for a 'real-time' framework. Please state clearly how this assumption affects the results or relax it in at least one sensitivity case.
  5. [Abstract and §1] The phrase 'for the first time' in the abstract and introduction is a strong novelty claim; since prior marginal emission factor work exists (e.g., Lusney 2020 and the references in the introduction), please temper the claim or explicitly differentiate the contribution.
  6. [§4] The conclusion states that the framework 'eliminates the need for compromised optimization with arbitrary tunable coefficients,' but the MEI estimation itself involves several discretionary choices (15 segments, choice of R² thresholds, emission factors, and the residual-demand model). Please rephrase to avoid overstating the parameter-free nature of the method.

Circularity Check

2 steps flagged · score 6.0 of 10

Reported emission reductions are the same fitted MEI used in the dispatch objective, so the benefit is an accounting consequence of the fitted slopes rather than an independent prediction.

  1. fitted input called prediction [Section 2.2, Eq. (5); Section 3.2, Table 2]
    "The objective function of ESS operation includes both electricity cost savings and carbon revenue, i.e.: max_{P^G_t} Σ_{t∈T} (λ^G_t + λ^C_t ρ^G_t) P^G_t. To quantitatively compare the performance of the ESS within and across the three cases, Table 2 report the numerical values for operational revenue and emission reduction."

    The only emission quantity defined in the paper is ρ^G_t from (2)–(3), whose slopes λ_{n,s} are fitted on six months of Ontario data. The optimization in (5) embeds λ^C_t ρ^G_t P^G_t as its carbon term, and Table 2 reports 'emission reduction' with no independent emissions model or out-of-sample benchmark. Hence the reported reduction is the same ρ^G-weighted net discharge used to drive the dispatch; any schedule that shifts P^G toward high-ρ hours automatically records a reduction. The prediction is forced by the fitted MEI rather than validated against an external emissions signal.

  2. self definitional [Section 2.1, Table 1]
    "the fifth column reports the total marginal supply response, calculated as the sum of the marginal slopes of all three resources. This value remains approximately equal to one across segments, verifying the proposed approach and that nearly the entire marginal demand is met by the combined contributions of gas, hydro, and imports."

    Residual demand is defined as total demand minus baseload/renewable generation, so gas + hydro + net imports is by energy balance the remaining supply that meets residual demand. Regressing each of these components on residual demand therefore yields slopes whose sum is approximately one by construction, regardless of whether any component is a causal marginal responder. Using this identity as 'verification' of the MEI approach is circular: it confirms the accounting relationship embedded in the regression setup, not the causal marginal-response interpretation needed for a real-time emission signal.

full rationale

The paper is not built on self-citation: the MEI coefficients λ_{n,s} are fitted from IESO data, emission factors are taken from external sources, and no uniqueness theorem or prior work by the same authors is load-bearing. The core circularity is in the numerical demonstration. Equation (5) optimizes dispatch using λ^C_t ρ^G_t, and Table 2's emission reduction is quantified with the same ρ^G_t; the paper provides no independent emissions model, out-of-sample evaluation, or external benchmark. The reported reduction is therefore algebraically tied to the fitted MEI, so the claimed benefit reduces to the fitted input. The sum-of-slopes check in Table 1 reinforces this: since the regressed resources approximately equal residual demand by energy balance, the near-unity sum is an accounting identity rather than causal validation. Overall, the MEI estimation itself has independent empirical content, but the paper's demonstration that the framework reduces emissions is circular, giving a partial-circularity score of 6.

Assumptions & free parameters 3 free parameters · 6 assumptions · 1 invented entities

The framework rests on a fitted MEI lookup table, assumptions about which generators are marginal, fixed emission factors, perfect foresight, and unvalidated acceptance of EPCs. The main burden is the regression-to-causality step: without independent validation of the slopes, the accounting is self-consistent but not verified.

free parameters (3)
  • Marginal supply shares lambda_n_s for 15 residual-demand segments = Table 1: gas -0.144 to 0.726; hydro 0.129 to 0.500; imports 0.098 to 0.880
    Fitted by cubic regression of each marginal resource output on residual demand, then piecewise-linearized into 15 segments; these slopes convert residual demand into MEI via Equation 2.
  • Segment boundaries for MEI lookup = -1000, 0, 1000, ... , 12000 MWh, 15 segments
    Chosen by hand in 1000 MWh increments; no sensitivity analysis is reported for this discretization.
  • MEI values m_s = Table 1: -0.053 to 0.369 tCO2e/MWh
    Derived as weighted sums of fitted slopes and emission factors, used directly in the dispatch objective in Equation 5; no confidence intervals are provided.
assumptions (6)
  • domain assumption Residual demand in Equation 1, total demand minus baseload and variable generation, isolates demand-driven changes in marginal resources.
    If non-dispatchable generation, especially hydro, is miscategorized, the fitted supply shares misattribute emissions. Footnote 2 acknowledges baseload hydro is not separated and promises an examination that does not appear.
  • domain assumption Gas, dispatchable hydro, and net imports are the only marginal units responding to residual demand; baseload and variable units are unresponsive.
    Section 2.1 states this based on IESO control logic; no empirical test of the identification assumption is provided.
  • domain assumption Emission factors are 0.37 tCO2e/MWh for gas, 0.44 tCO2e/MWh for imports, and zero for hydro.
    The imports factor is a single fixed value regardless of time, source, or neighboring grid mix, cited from Electricity Maps 2025; misestimation changes MEI directly.
  • domain assumption The ESS has perfect foresight of electricity price, carbon price, and MEI over the optimization horizon.
    Section 3.1 states dispatch is based on accurate predictions; no forecasting error model is included, which overstates achievable emission reductions and revenue.
  • ad hoc to paper Emission performance credits are accepted by compliance carbon markets as tradable offset instruments.
    The framework assumes the proposed credits are recognized by regulated entities; no regulatory or market validation is provided.
  • standard math The dispatch optimization is a linear program implemented in Pyomo and solved with Gurobi.
    Standard mathematical programming assumption; the formulation in Section 2.2 is internally coherent.
invented entities (1)
  • Emission performance credits (EPCs)
    purpose: Tradable credits representing avoided emissions from ESS operation, sold to regulated entities to meet compliance targets.
    No regulator or market has adopted EPCs; the paper provides no external verification, additionality test, or baseline standard. It is a proposed accounting instrument, not an observed entity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Emission-Aware Operation of Electrical Energy Storage Systems." pith.science (2026). https://pith.science/paper/YECW5H57

@misc{pith2026250616454,
  author       = {Pith},
  title        = {Pith review of: Emission-Aware Operation of Electrical Energy Storage Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YECW5H57}},
  note         = {Machine review of arXiv:2506.16454}
}
read the original abstract

Since the beginning of this century, there has been a growing body of research and developments supporting the participation of energy storage systems (ESS) in the emission reduction mandates. However, regardless of these efforts and despite the need for an accelerated energy transition, we have yet to see a practical framework for operational carbon accounting and credit trading for energy storage systems. In this context, this paper proposes an emission performance credits (EPCs) framework that allows ESS, down to the prosumer level, to participate in the carbon market. Thus, a mechanism is proposed, for the first time, to calculate the grid's real-time marginal emission intensity (MEI). The MEI is then used to optimize the cumulative operational emission of ESS through carbon-aware dispatch. Consequently, the framework tracks the operational emissions and converts them into EPCs, which are then sold to regulated entities under compliance programs. Simulation results support the potential of ESS, regardless of their size, to participate in the broader carbon mitigation objectives.

Figures

Figures reproduced from arXiv: 2506.16454 by the authors.

Figure 1
Figure 1. Marginal units output vs. residual demand. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. ESS carbon-aware dispatch structure. determines the optimal ESS dispatch. The objective function of ESS operation includes both electricity cost savings and carbon revenue, i.e.: max P G t X t∈T [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Hourly residual demand and its duration curve [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Hourly electricity price and its duration curve. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: ESS dispatch across three pricing cases with different ESS capacity. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Normalized average performance (wider better). Together, these figures highlight the importance of coordinating fully cost-reflective pricing strategies and ESS capacity to achieve an economical decarbonization. 4. Conclusion and Future Works This paper presents an emi…
Figure 7
Figure 7. Figure 7: Impact of ESS capacity and carbon price on [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

3 extracted references · 3 canonical work pages

  1. [1]

    M., El-Taweel, N

    Abomazid, A. M., El-Taweel, N. A., & Farag, H. E. Z. (2022). Optimal energy management of hydrogen energy facility using integrated battery energy storage and solar photovoltaic systems. IEEE Transactions on Sustainable Energy, 13(3), 1457–1468. Azuatalam, D., F ¨orstl, M., Paridaric, K., Ma, Y ., Chapman, A. C., & Verbi ˇc, G. (2018). Techno-economic ana...

    work page 2022

  2. [2018]

    Babacan, O., Abdulla, A., Hanna, R., Kleissl, J., & Victor, D. G. (2018). Unintended effects of residential energy storage on emissions from the electric power system. Environmental science & technology, 52(22), 13600–13608. Bhattacharjee, S., Sioshansi, R., & Zareipour, H. (2021). Benefits of strategically sizing wind-integrated energy storage and transm...

    work page 2018

  3. [4592]

    Gao, H., Jin, T., Wang, G., Chen, Q., Kang, C., & Zhu, J. (2024). Low-carbon dispatching for virtual power plant with aggregated distributed energy storage considering spatiotemporal distribution of cleanness value. Journal of Modern Power Systems and Clean Energy, 12(2), 346–358. Government of Alberta. (2023). Technology innovation and emissions reductio...

    work page Pith review arXiv 2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.