Pith. sign in

REVIEW 5 major objections 6 minor 25 references

Optimal Co-Design of a Hybrid Energy Storage System for Truck Charging

T0 review · 5 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read For a truck-charging microgrid, hybrid battery-plus-fast storage cuts 20-year cost by 1.96% versus battery-only.

desk verdict Useful co-design framework for hybrid storage in truck-charging microgrids, but the headline result that the fully hybrid solution is best is not supported by the reported data—Exp2 and Exp4 are tied within solver tolerance and the text contradicts Table 2. read the letter →

arxiv 2506.01426 v1 pith:UEKN5E3J submitted 2025-06-02 eess.SY cs.SY

Juan Pablo Bertucci , Sudarshan Raghuraman , Mauro Salazar , Theo Hofman This is my paper
classification eess.SYcs.SY
keywords hybridenergystorageco-designmixed-integerlinearprogrammingmicrogridbatteryelectrictruckchargingsupercapacitorflywheelnetpresentvalue
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 argues that stationary storage for battery-electric truck charging should be designed and operated in a single optimization, because sizing and daily dispatch decisions are coupled. For a distribution-center microgrid in the Netherlands, the authors claim that a battery-only system is already competitive, but that adding supercapacitors and a flywheel lowers the 20-year total cost by 1.96% at a 2.6% higher initial investment. The result comes from solving a mixed-integer linear program with global optimality guarantees, using a synthetic representative month built from three years of solar, price, and charging data. A sympathetic reader would take the paper's core message to be that hybrid storage can pay for itself through lower operating cost and reduced grid dependence, without requiring a heuristic or sequential design process.

What carries the argument

The central object is a mixed-integer linear program in which installed capacities of the battery, supercapacitor, flywheel, solar panels, and grid connection are optimized together with every hourly power dispatch decision. The formulation stays linear by introducing an auxiliary variable $q_{e,k}=E^{\max}_e R_{e,k}$ for storage throughput, so the C-rate limit $R_{e,k} \le R^M_e$ can be enforced without bilinear products; the paper states that this convex relaxation is lossless for selling factors in $(0,1]$. The optimization horizon is a synthetic 30-day month assembled by k-means clustering of historical days and a Markov chain over cluster transitions, with costs annualized over 20 years at a discount rate.

What would settle it

Re-run the identical objective on the full three years of historical hourly data instead of the synthetic 30-day month: if the optimal storage mix changes or the hybrid advantage over battery-only disappears, the central claim fails. A sharper test is to solve one representative week at one-minute resolution; if the selected supercapacitor and flywheel capacities change substantially at that resolution, the hourly co-design was missing the very transients that justify them.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is that the fully hybrid system—battery, supercapacitor, and flywheel together—is the optimal solution of a jointly optimized design-and-control problem for the studied truck-charging microgrid. It achieves the lowest total cost of ownership, about 1.96% lower than battery-only, and the lowest operating cost, while keeping energy sales to the grid nearly unchanged and slightly reducing purchased energy. The authors also show that battery-only has the lowest capital cost, making the choice a trade-off between upfront investment and long-term operating expenses. They frame this as evidence that co-design, rather than sequential sizing followed by scheduling, can reveal storage combinations that a battery-only design would miss.

Load-bearing premise

The numbers depend on a synthetic 30-day month, generated by clustering three years of data and a Markov chain, being representative of 20 years of operation at an hourly resolution, including the fast power transients that supercapacitors and flywheels exist to handle; if that month misrepresents those transients, the chosen sizes and the 1.96% saving are unreliable.

Editorial extensions

If this is right

  • Sites that can accept a 2.6% higher capital outlay can expect lower 20-year total cost and lower operating expenses from the fully hybrid storage system.
  • Adding supercapacitors alone or a flywheel alone each reduce total cost relative to battery-only, so partial hybridization is also a rational intermediate choice.
  • Because the MILP is solved with global optimality guarantees, the cost ranking among storage configurations is not an artifact of heuristic sizing or scheduling.
  • The framework transfers to other sites by replacing solar, price, and charging-load inputs, making the choice of storage mix a data-driven calculation rather than a rule of thumb.

Reading between the lines

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

  • The 1.96% saving is about the same size as typical uncertainty in storage capital costs, so under different price assumptions the ranking between hybrid and battery-only could reverse; the paper itself flags sensitivity to price assumptions.
  • An hourly time step may undervalue supercapacitors and flywheels, whose main benefit is sub-hourly response; a one-minute-resolution study of the same site could either strengthen the hybrid case or show that hourly operation is what actually drives cost.
  • The model sells energy back to the grid but does not price ancillary services such as frequency regulation; adding those revenue streams would likely improve the economics of fast storage further.
  • Because the charging profiles come from schedules optimized to limit peak consumption, the case study may be unusually friendly to battery-only storage; uncoordinated charging elsewhere could increase the measured benefit of fast storage.
Share X Bluesky LinkedIn Reddit HN

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. This paper proposes a co-design framework for a microgrid serving battery-electric truck charging, in which the capacities of battery, supercapacitor, flywheel, PV, and grid connection are optimized together with hourly dispatch over a 30-day synthetic period. Four storage configurations are compared in a Dutch distribution-center case study, and the authors report that the fully hybrid configuration achieves the lowest total cost, 1.96% below the battery-only baseline, at a higher capital cost. The framework is formulated as a linearized optimization problem that the authors call a MILP, with parameters from public sources and charging loads from previous work.

Significance. The topic is relevant and timely, and the paper has some strengths: the model couples sizing and operation, parameter values are taken from cited external reports, and the optimization pipeline (YALMIP/Gurobi) is standard. The synthetic-scenario construction via k-means and a Markov chain is clearly described. However, the numerical results as reported do not support the central claim: the best-configuration ranking is within the solver optimality gap, the tables and prose contradict each other, and the 30-day/hourly modeling choices cannot capture the sub-hourly transients that motivate supercapacitors and flywheels. Until these issues are resolved, the claimed cost saving is not established.

major comments (5)
  1. [Table 2 and Abstract/§3] The claim that Experiment 4 achieves the lowest total cost is not supported by the reported data. Table 2 lists total costs of 22.387 for Exp2 and 22.386 for Exp4, a difference of 0.001 units, while the text reports a solver gap below 1%; a 1% gap on a cost of about 22.4 units is about 0.224 units, so the optimizer cannot resolve a 0.001-unit difference. The 1.96% saving relative to Exp1 is of the same order as the gap and therefore also not established.
  2. [Section 3 vs Table 2] The reported results are internally inconsistent. For Exp4, the text gives CapEx 2.520 and OpEx 20.287, whereas Table 2 gives CapEx 2.629 and OpEx 19.757. Moreover, using Table 2's own entries, CapEx + OpEx - EOL = 2.629 + 19.757 - 0.422 = 21.964, not the tabulated total cost 22.386; the text's numbers (2.520 + 20.287 - 0.422 = 22.385) come closer. The abstract's 2.6% higher initial investment also matches Table 2, not the text's 1.64%. These contradictions mean the supporting numbers are not reproducible.
  3. [Section 2.6, Eq. (31), Table 1] The optimization is performed over a 30-day synthetic period, yet the objective is presented as a 20-year net present value. Equation (31) sums discounted yearly costs, but the paper never specifies how the 30-day simulation cost is annualized (e.g., by a factor of 365/30) or validated against annual totals, seasonal variability, or peak events. The extrapolation from one synthetic month to 20 years is therefore undescribed, and the resulting 1.96% saving cannot be interpreted as a long-run cost reduction.
  4. [Section 2.3, Table 1, Fig. 3] With τ=60 min, the model constrains only hourly average power and hourly energy transitions. The paper motivates supercapacitors and flywheels by their ability to handle rapid, sub-hourly fluctuations, but an hourly discretization cannot represent such transients; the C-rate constraints in (18) limit changes per hour, not instantaneous power. As a result, the cost differences among configurations with and without supercapacitors/flywheels are not anchored to the physical capability that these technologies are supposed to provide.
  5. [Section 2.5, Problem 1] The problem is called a mixed-integer linear program, but no integer or binary decision variables appear in the formulation; all variables in Problem 1 are stated to be real, and the k-means cluster assignments in Section 2.6 are pre-processing, not part of the optimization. The problem is therefore a linear program after the McCormick linearization, and the term 'MILP' is inaccurate.
minor comments (6)
  1. [Section 3, after Table 2] 'sold energy and purchased energy increasing slightly with Exp. 2' should refer to Exp. 3; Table 2 shows Exp3 sold and purchased energies higher than Exp2.
  2. [Conclusions] 'adding flywheels (Exp. 3 and Exp. 4) increases sold energy' is contradicted by Table 2, where Exp4 sold energy (234.53) is slightly below Exp1 (234.73).
  3. [Eq. (10)] The constraint has the quantifier '∀s ∈ S' but applies to the grid; it should be '∀k ∈ K' or explicitly to grid power.
  4. [Eq. (36)] The formula places J_cap_s inside the same parenthesis as the storage resale term; please clarify the intended expression and check the dimensions of the cycle-based resale value.
  5. [Table 1] The table contains apparent typos: the second 'CPe' entry (0.001) is likely a different parameter, and 'Cop' is undefined; units are inconsistently written as 'k e' and 'ke'.
  6. [Throughout] There are numerous spelling errors (e.g., 'repsented', 'agreggation', 'Mantainance', 'pondered', 'disount', 'callibrated', 'exsting', 'comparsion', 'vehcles'); the paper needs a careful proofread.

Circularity Check

0 steps flagged · score 1.0 of 10

No circularity: the co-design result is a direct optimization output; the only self-citation supplies the charging-load input, not the cost comparison.

full rationale

The optimization chain is self-contained: Problem 1 with Eqs. (1)-(37) takes externally sourced PV, grid-price, and equipment-cost parameters, plus a synthetic demand sequence, and produces optimal storage sizes and dispatch by solving a MILP with Gurobi to a 1% gap. The claimed 1.96% saving is read directly from the objective values in Table 2 (22.834 k€ for Exp. 1 versus 22.386/22.387 k€ for Exp. 2 and Exp. 4) and is not a parameter fitted to reproduce the conclusion. The only self-citation, Bertucci et al. (2024), is used in Section 3 as the source of the EV charging-load input Pd; the paper's central claim about which storage mix minimizes cost does not reduce to that citation. The numerical inconsistencies between the prose and Table 2 (e.g., Exp. 4 CapEx quoted as 2.520 k€ versus 2.629 k€ in the table, and OpEx quoted as 20.287 k€ versus 19.757 k€) are correctness and reproducibility concerns, not circularity. Similarly, the near-tie between Exp. 2 and Exp. 4 within the reported 1% solver gap, while it weakens the 'fully hybrid is best' claim, is a support issue rather than a self-referential derivation. The authors' closing caveat that results are sensitive to price assumptions and that charging powers were optimized to minimize peak consumption is an honest limitation, not a circular step.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central result rests mainly on scenario reduction choices (W, Tsyn, tau) and the unverified relaxation claim, rather than on fitted model parameters. The parameters from the literature are inputs, not free parameters, but their uncertainty is not propagated.

free parameters (5)
  • Number of clusters W = 20
    Chosen by authors for k-means scenario reduction; the resulting 30-day synthetic month drives all sizing and cost results.
  • Synthetic period length Tsyn = 30 days
    The optimization horizon is one synthetic month, which is then extrapolated to 20 years; representativeness is not validated.
  • Time step tau = 60 min
    Hourly discretization cannot resolve sub-hourly supercapacitor and flywheel transients that the paper claims these devices handle.
  • Sell factor fsell = 0.75
    Assumed fraction of grid price received for exported energy; no sensitivity analysis is provided.
  • Depth of discharge bounds E_min = [0.15, 0, 0] (fraction)
    Battery DoD assumed 15% lower bound; other storages 0; these bounds shape the usable capacity and hence sizing.
assumptions (5)
  • domain assumption Hourly time resolution is adequate to capture the operational differences among battery, supercapacitor, and flywheel storage.
    Used throughout the MILP formulation and Fig. 3; load-bearing because supercapacitors and flywheels are justified by fast transients.
  • domain assumption The 30-day synthetic month generated by k-means and a Markov chain represents the 20-year operational period for cost NPV.
    Section 2.6 constructs Tsyn but never describes how the 30-day result is scaled to Y=20 years in Eq. (31).
  • ad hoc to paper The convex relaxation of the complementarity between positive and negative power flows is lossless for fsell in (0,1].
    Stated in Section 2.5 without proof or reference; it is what lets the problem be solved as an LP rather than a true MILP.
  • ad hoc to paper McCormick envelope for q_e,k = Emax_e * R_e,k is exact at the optimum.
    Eqs. (19)-(24) relax the bilinear term; exactness relies on the cost minimizing q_e,k, which is plausible but not shown.
  • domain assumption The battery electric truck charging demand from Bertucci et al. (2024) and partner data is representative for this distribution center.
    Used as the main demand input; authors note charging powers may differ in other plazas.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optimal Co-Design of a Hybrid Energy Storage System for Truck Charging." pith.science (2026). https://pith.science/paper/UEKN5E3J

@misc{pith2026250601426,
  author       = {Pith},
  title        = {Pith review of: Optimal Co-Design of a Hybrid Energy Storage System for Truck Charging},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UEKN5E3J}},
  note         = {Machine review of arXiv:2506.01426}
}
read the original abstract

The major challenges to battery electric truck adoption are their high cost and grid congestion.In this context, stationary energy storage systems can help mitigate both issues. Since their design and operation are strongly coupled, to make the best out of them, they should be jointly optimized. This paper presents a co-design framework for hybrid energy storage systems where their technology and sizing are optimized jointly with their operational strategies. Specifically, we consider a microgrid supporting truck chargers that consists of utility grid, solar panels, and energy storage systems including batteries, supercapacitors and flywheels. We frame the co-design problem as a mixed-integer linear program that can be solved with global optimality guarantees. We showcase our framework in a case-study of a distribution center in the Netherlands. Our results show that although the battery-only configuration is already competitive, adding supercapacitors or flywheel storage decrease total cost and increase energy sold back to the grid. Overall, the fully hybrid solution (Battery+Supercapacitors+Flywheel) offers the best outcomes, achieving the lowest overall cost (1.96\% lower compared to battery-only) and reduced grid dependency, but at a higher (2.6\%) initial investment.

Figures

Figures reproduced from arXiv: 2506.01426 by the authors.

Figure 1
Figure 1. Sketch of the microgrid and main problem variables. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Obtained cluster groups for grid prices CG, power demand Pd and solar power production PPV,with each colored dot representing a different cluster Design Results [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Power Demands [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

25 extracted references · 23 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address author booktitle chapter doi edition editor eid howpublished institution journal key month note number organization pages publisher school series title type url volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.all := #1 'mid.sent...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Bertucci, J.P., Hofman, T., and Salazar, M. (2024). Joint optimization of charging infrastructure placement and operational schedules for a fleet of battery electric trucks. In 2024 American Control Conference, 2995--3000

    work page 2024

  4. [4]

    Bian, M., Xu, Z., and Li, F. (2019). Design and control of a hybrid microgrid with supercapacitor-based energy storage system for peak load shaving. Applied Energy, 239, 1638--1647

    work page 2019

  5. [5]

    Chong, L., Wong, Y., Rajkumar, R., and Isa, D. (2016). Hybrid energy storage systems and control strategies for stand-alone renewable energy power systems. Renewable and Sustainable Energy Reviews, 66, 174--189

    work page 2016

  6. [6]

    Ding, H., Hu, Z., and Song, Y. (2015). Value of the energy storage system in an electric bus fast charging station. Applied Energy, 157, 630--639

    work page 2015

  7. [7]

    Photovoltaic geographical information system (pvgis)

    ECJRC (2025). Photovoltaic geographical information system (pvgis). ://joint-research-centre.ec.europa.eu/

    work page 2025

  8. [8]

    Transparency platform

    ENTSO-E (2025). Transparency platform. ://transparency.entsoe.eu/

    work page 2025

Show all 25 references
  1. [9]

    Etukudoh, E., Adefemi, A., Ilojianya, V., Umoh, A., Ibekwe, K., and Nwokediegwu, Z. (2024). A review of sustainable transportation solutions: Innovations, challenges, and future directions. World Journal of Advanced Research and Reviews

    2024

  2. [10]

    Gurobi Optimizer Reference Manual

    Gurobi Optimization, LLC (2025). Gurobi Optimizer Reference Manual . ://www.gurobi.com

    2025

  3. [11]

    Renewable power generation costs in 2023

    IRENA (2024). Renewable power generation costs in 2023

    2024

  4. [12]

    Kebede, A.A., Kalogiannis, T., Van Mierlo, J., and Berecibar, M. (2022). A comprehensive review of stationary energy storage devices for large scale renewable energy sources grid integration. Renewable and Sustainable Energy Reviews, 159, 112213

    2022

  5. [13]

    Krein, P. (2020). Flywheel energy storage for frequency regulation: A survey. IEEE Transactions on Power Electronics, 35(2), 1631--1639

    2020

  6. [14]

    Li, D.G., Zou, Y., Hu, X.S., and Sun, F.c. (2012). Optimal sizing and control strategy design for heavy hybrid electric truck. In IEEE Vehicle Power and Propulsion Conference, 1100--1106

    2012

  7. [15]

    Lin, C.C., Peng, H., Grizzle, J., and Kang, J.M. (2003). Power management strategy for a parallel hybrid electric truck. IEEE Transactions on Control Systems Technology, 11(6), 839--849

    2003

  8. [16]

    Liu, G., Xue, Y., Chinthavali, M., and Tomsovic, K. (2020). Optimal sizing of PV and energy storage in an electric vehicle extreme fast charging station. In IEEE Power and Energy Society Innovative Smart Grid Technologies Conference, 1--5

    2020

  9. [17]

    L \" o fberg, J. (2004). Yalmip: A toolbox for modeling and optimization in matlab. In Proceedings of the CACSD Conference. Taipei, Taiwan

    2004

  10. [18]

    Minear, E., Ip, P., Westlake, B., Alexander, M., Evans, M., Pellow, M., and Kamath, H. (2018). Energy storage technology and cost assessment: Executive summary. Technical report, Technical Report 3002013958, Electric Power Research Institute

    2018

  11. [19]

    Mongird, K., Viswanathan, V.V., Balducci, P.J., Alam, M.J.E., Fotedar, V., Koritarov, V., and Hadjerioua, B. (2019). Energy storage technology and cost characterization report. Technical report, Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)

    2019

  12. [20]

    Ortega, E.M., Rodriguez, L., and Martin, F. (2018). Analysis of flywheel energy storage performance for microgrid frequency regulation: Techno-economic benefits. Energy, 145, 120--130

    2018

  13. [21]

    and Venkataramanan, G

    Smith, J. and Venkataramanan, G. (2016). An overview of flywheel energy storage systems in microgrids. IEEE Power and Energy Magazine, 14(5), 30--39

    2016

  14. [22]

    and Park, S

    Song, F.U. and Park, S. (2020). Integration of supercapacitor energy storage for frequency regulation in islanded microgrids. IEEE Transactions on Industrial Electronics, 67(6), 4853--4864

    2020

  15. [23]

    Addressing the heavy-duty climate problem

    Transport & Environment (2022). Addressing the heavy-duty climate problem

    2022

  16. [24]

    and Mateescu, V

    Tudorache, T. and Mateescu, V. (2019). Flywheel energy storage systems for frequency regulation in stand-alone microgrids. IEEE Transactions on Industrial Electronics, 66(5), 3942--3953

    2019

  17. [25]

    Zhang, X., Li, Y., and Qiu, J. (2022). Optimal sizing and scheduling of supercapacitor storage in microgrids for cost reduction and reliability improvement. Journal of Energy Storage, 46, 10351--10361

    2022

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.