Quantitative Indices for Improving Metro Load Curve, Using Distributed Generation

Alireza Fereidunian; Masoud Behbahani

REVIEW 3 major objections 7 minor 16 references

Quantitative Indices for Improving Metro Load Curve, Using Distributed Generation

T0 review · 3 major / 7 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 The paper claims that installing distributed generation sized to 40% of a metro's peak demand raises its load factor from 0.53 to 0.73, while cutting peak grid demand and transmission losses.

desk verdict A well-organized concept note on metro peak shaving with DG, but its quantitative indices rest on an unvalidated load curve and a load-factor calculation that does not add up from the paper's own assumptions. read the letter →

arxiv 2506.08975 v1 pith:5GDEE555 submitted 2025-06-10 eess.SY cs.SY

Masoud Behbahani , Alireza Fereidunian This is my paper

classification eess.SYcs.SY

keywords loadcurvedistributedgenerationmetrorailwayfactorpeakshavingtractionpowersubstationlightanddemandsidemanagement

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 a metro railway's daily demand curve, two rush-hour peaks and a low midday valley, can be substantially improved by installing distributed generation that runs only during those peaks. Starting from a typical curve built by scaling traction load to passenger counts and light-and-power load to 60% of traction, the authors calculate a load factor of 0.53. Sizing the generators as peak minus midday base demand gives 40% of peak, and the resulting curve has a load factor of 0.73. If this pattern holds, DG becomes an attractive load-curve improvement for metros, since the same units also replace emergency diesel generators and reduce battery backup.

What carries the argument

The central objects are the 24-hour metro load curve and the load factor $LF=P_{avg}/P_{max}$. The carrying identity is the sizing rule $P_{DG}=P_{peak}-P_{base}$; with $P_{base}=0.6P_{peak}$ (from an assumed 60% LPS-to-TPS ratio) this fixes $P_{DG}=0.4P_{peak}$. The load-curve extraction converts passenger timetables and substation load proportions into a curve on which these indices can be computed before any feeder measurement is made.

What would settle it

Instrument the outgoing feeders of a metro's traction and light-power substations for one working day, build the actual 24-hour curve, and recompute the load factor and peak-minus-base DG size; if the measured base demand is not near 60% of peak, or the measured load factor departs materially from 0.53, the claimed 40% sizing and 0.73 result will not hold.

Watch

Extended reading notes

Core claim

The paper's central claim is that a metro's doubly peaked load curve can be flattened, in quantitative index terms, by a simple DG sizing rule: let the distributed generators supply exactly the difference between the peak demand and the midday base demand. On the paper's typical curve this means $P_{DG}=P_{peak}-P_{base}=0.4P_{peak}$, and the load factor rises from 0.53 to 0.73 while grid demand during peaks falls by 40% and peak-time transmission losses fall by 64%. The same DG units can serve as the required emergency backup, turning a cost into an asset.

Load-bearing premise

The quantitative result rests on an assumed, not measured, daily load curve: traction demand proportional to passenger counts, and light-and-power demand fixed at 60% of traction demand.

Editorial extensions

If this is right

A metro using this DG sizing draws 40% less energy from the grid during rush hours, lowering demand charges and cutting peak-time transmission losses by 64%.
The load factor rises from 0.53 to 0.73, meaning the same substation and line capacity delivers more useful average energy per unit of installed capacity.
Emergency diesel generators become redundant for peak coverage, and battery/UPS support time can be shortened, saving capital and maintenance cost.
Because metro peaks overlap national peaks, shaving them relieves the grid's most expensive peak generation at the hours when it would otherwise run.
The same quantitative indices let an operator estimate initial investment and return on investment before choosing DG over other load-curve methods.

Reading between the lines

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

If the peak-minus-base rule is sound, the exact DG fraction of 40% is an artifact of the assumed 60% LPS-to-TPS ratio; measured feeder data could shift it, so the method generalizes even if this specific number does not.
The approach could transfer to monorails, trolleybuses, or other transit loads with a midday valley, but each needs its own measured curve before the claimed indices apply.
Because the post-DG curve still has a midday valley, adding distributed storage or valley-filling DSM could raise the load factor beyond 0.73, a combination the paper does not explore.
The economic payoff depends on demand tariffs and peak-energy pricing; in flat-tariff regions the 0.73 load factor improves asset utilization without necessarily cutting the energy bill.

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.

Desk Editor's Note

read the letter

Thanks for sharing the report. I read the paper alongside the stress-test note, and I largely agree with the rejection, though I would put the emphasis slightly differently.

The paper's virtue is that it is clearly organized and addresses a real problem: metro load curves have two peaks that coincide with grid peaks, and because the metro is fed by a dedicated network, DG is technically a feasible way to shave those peaks. The decomposition into traction (TPS) and light and power (LPS) loads with a 60/40-ish split is a reasonable engineering heuristic, and the sizing rule PDG = Ppeak - Pbase is a sensible starting point for a feasibility study. For an engineer wanting a back-of-the-envelope method, that part is useful.

The soft spots are usual ones, and they are load-bearing. The 'typical load curve' that drives all the numbers is not measured; it is assembled from an assumed 60% LPS-to-TPS ratio and a train-count curve. The paper itself concedes that actual feeder measurements should be the criterion in practice. That concession is honest, but it undercuts the specific indices: the baseline LF of 0.53 depends entirely on the assumed curve.

More importantly, the paper's flagship number—LF improving from 0.53 to 0.73—does not follow from the stated assumptions. The paper says the DG runs from 7 to 9 am and 5 to 7 pm, four hours total, and sizes it at 0.4 Ppeak, which implies 1.6 Ppeak-h of removed energy under the favorable assumption that the load sits at peak the whole time. With a post-DG peak of 0.6 Ppeak, the LF should be about 0.77, not 0.73. To hit 0.73 you would need either a longer operating window or a higher post-DG peak, and neither is stated. The stress-test arithmetic checks out. So the central quantitative claim is not reproducible from the paper's own equations, and no sensitivity analysis or real data is offered.

The 64% transmission-loss-reduction bullet is just a citation to a 1999 conference paper, so it does not carry the weight the discussion gives it.

This is a plausible concept note, not a research result. For a trade magazine or a project pre-feasibility document, it would be fine. As a peer-reviewed contribution, it does not meet the bar: the quantitative conclusions are unvalidated and, as written, internally inconsistent. I would not send this to a serious referee, and I would not cite the LF numbers for any real planning.

Referee Report

3 major / 7 minor

Summary. The paper proposes using distributed generation (DG) to improve the load curve of an urban metro system. It constructs a "typical" daily metro load curve from assumed proportions between traction (TPS) and light/power (LPS) substations, computes a baseline load factor of 0.53, sizes a DG at 40% of peak load using PDG = Ppeak - Pbase, and claims the load factor improves to 0.73, with 40% demand reduction and 64% transmission-loss reduction. The authors argue that metro systems have a particularly attractive techno-economic case for DG because of their dedicated network, heavy demand, and poor load factor.

Significance. If the quantitative indices were reproducible and based on measured feeder data, the paper would offer a useful preliminary screening tool for metro operators considering DG. The paper's strengths are its clear definition of load factor, the explicit DG sizing formula, the review of DSM/DS/DG options, and an explicit admission that actual feeder measurements should be the criterion in practice. However, the central numerical results are not derived from validated data or from a fully specified calculation, and therefore the specific quantitative claims cannot be accepted as presented.

major comments (3)

[Section III-B, Fig. 7] The baseline load curve and the resulting load factor 0.53 are built from assumed proportions rather than measured feeder data: the text states "LPS power requirements is between 50% to 70% of the TPS, so assuming a 60% combined ratio..." and later concedes that "in a practical project measures taken outlet feeders can be a criterion." Because every downstream number (PDG = 0.4 Ppeak, LF improvement to 0.73, demand and loss reductions) is derived from this assumed curve, the quantitative claims are unvalidated and would change with any real metro load shape.
[Section IV-A/IV-B] The calculation producing LF = 0.73 is not shown and does not follow from the stated assumptions. Equation (2) sizes PDG = Ppeak - Pbase = 0.4 Ppeak, and Section IV-B says DG operates only from 7-9 am and 5-7 pm (four hours). Under the favorable assumption that all demand above 0.6 Ppeak falls in those windows, the after-DG load factor would be (0.53 - (0.4 x 4)/24)/0.6 ≈ 0.77, not 0.73. To obtain 0.73 the DG must supply roughly 2.2 Ppeak·h, which corresponds to about 5.5 hours at rated power, or the net peak after DG must remain above 0.6 Ppeak. The paper supplies neither the energy balance nor the post-DG average and peak values needed to reproduce the headline number.
[Section IV-B] The bullet "A 64% reduction in transmission losses at peak times [16]" is presented as a quantitative result of this study, but it is taken from an external rule of thumb and not derived for the proposed DG dispatch or the metro network. Similarly, "A 40% reduction in electricity demand" is ambiguous: it could mean a 40% reduction in peak demand, but the actual energy supplied by a 0.4 Ppeak DG running four hours is only about (0.4 x 4)/24 = 6.7% of daily energy, so the claim needs a clear definition of "demand" and a supporting calculation.

minor comments (7)

[Eq. (1)] Define "Average Load" and "Maximum Demand" over the same time interval (e.g., daily) to make LF unambiguous.
[Section III-B] Specify "60% combined ratio" as the LPS-to-TPS power ratio; Fig. 7 axes are in percent with no absolute power scale, making the LF calculation hard to verify.
[Section IV-A and IV-B] The Pbase interval is 9:30 am-4 pm, while the DG window is 7-9 am and 5-7 pm; the relationship between these intervals and the shape of the assumed curve should be explained.
[Section II-C] The sentence "Although Metro is runs around the city..." is incomplete; add the missing clause.
[References] Reference [3] cites a blog post; replace it with a primary data source for London station passenger flows.
[Abstract] The abstract states "This paper promises the idea" and later "Distribute Generation"; proofread for language and typos.
[Fig. 8] Describe how the "after DG" curve is obtained (which hours are curtailed and by how much) to allow visual verification of the 0.73 LF.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the DG sizing rule and LF figures are scenario calculations from an explicitly assumed load curve, not predictions forced by definition or by self-citation.

full rationale

The paper's derivation chain is: assume a typical Metro load curve (TPS proportional to passenger count, LPS taken as 60% of TPS), compute the baseline LF = 0.53 from that curve, define a DG capacity PDG = Ppeak - Pbase where Pbase is the off-peak average (0.6Ppeak for this curve), then simulate the effect of DG. This is a self-consistent model exercise: the output LF is a function of the input curve and the DG dispatch, not identical to any input by construction. The paper itself concedes in Section III-B that 'The calculated LF is based on the assumptions considered for simplification of calculation; in a practical project measures taken outlet feeders can be a criterion to determine the load curve.' That is an unvalidated assumption, not circularity. The claim that LF improves from 53% to 73% is not explicitly derived in the equations, and the stated 4-hour DG operating window (7-9 am, 5-7 pm) would under favorable assumptions give approximately 0.77, not 0.73; this is a reproducibility/correctness concern, not a circularity. The only self-citation ([11], a textbook on induction motor control) is used for fleet-efficiency technology and is not load-bearing for the DG/load-factor claim. No equation is defined in terms of the result it is said to predict, and no fitted parameter is renamed as a prediction. Therefore no significant circularity is found.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

All quantitative indices rest on an assumed load curve and a chosen sizing rule; no empirical data, sensitivity analysis, or independent validation is provided.

free parameters (4)

Pbase ratio (60% of Ppeak) = 0.6
Chosen from the assumed load curve and the stated LPS/TPS range of 50-70%; used to size DG at 40% of peak (Eq. 3).
LPS to TPS consumption ratio = 0.6
Assumed in Section III-B as 60% within the 50-70% range, without measurement.
DG operating windows = 7-9 am and 5-7 pm
Chosen to match assumed morning and evening peaks; no real timetable data provided.
Typical load curve shapes (Figs. 5-7)
Constructed from assumptions about passenger traffic and train timetables; not based on recorded feeder data.

assumptions (4)

domain assumption Metro load is proportional to passenger count and train headway
Section III-A states the TPS load curve is proportional to the number of running trains/passengers without empirical verification.
domain assumption The synthesized curve of Fig. 7 represents a typical metro load curve
The paper uses this curve to compute LF=0.53; no real measurements are provided.
domain assumption DG can be connected in parallel to the metro's dedicated supply network
Section II-C argues feasibility from the exclusive network, but no technical integration study is given.
domain assumption Peak-time transmission losses are 64% of total losses
Cited from reference [16] without derivation; used to claim 64% loss reduction in Section IV-B.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Quantitative Indices for Improving Metro Load Curve, Using Distributed Generation." pith.science (2026). https://pith.science/paper/5GDEE555

@misc{pith2026250608975,
  author       = {Pith},
  title        = {Pith review of: Quantitative Indices for Improving Metro Load Curve, Using Distributed Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5GDEE555}},
  note         = {Machine review of arXiv:2506.08975}
}

read the original abstract

This paper promises the idea of using DG (Distributed Generation) to improve the Metro load curve. Public transportation systems are often based on gasoline and diesel. However, with the gradual development in usage of the Metro and monorail, a new load with heavy demand, inappropriate load curve and middle LF (Load factor) is added to the electricity grid. In addition to supply problem of this massive consumer, the Metro load curve is another problem, which has a relatively low LF. Furthermore, Metro load peak hours coincide with the peaks of national grid. Improvement of the load curve is well-known in electrical engineering literature, which depending on the type of load curve, offers general recommendations in three approaches; DSM (Demand Side Management), DS (Distributed Storage) and DG. In this paper, to achieve quantitative indices of improvement for Metro load curve using DG, firstly based on the analysis of volume and consumption pattern of the main loads in Metro, the typical load curve has been extracted. Using this curve, the result of using DG is shown by quantitative parameters which represent the significant improvement in load curve. These parameters can be used to calculate economic indicators such as initial cost and ROI (Return of Investment).

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

16 extracted references · 16 canonical work pages

[16]

High proportion of peak load losses in the total electrical energy losses

A. Sharifi Yazdi, “High proportion of peak load losses in the total electrical energy losses", Second National Energy Conference , (1999)

work page 1999
[1]

Energy Balance Report 2010

Iranian Ministry of Energy, Department of Power and Energy, "Energy Balance Report 2010", (2010)

work page 2010
[2]

Forecasting Electrical energy demand of the transport sector of Iran using MAED model by the year 2016

O. Shah Hoseini, Z. Soleimanian, R. Goodarzi and M. Alimardani, "Forecasting Electrical energy demand of the transport sector of Iran using MAED model by the year 2016", Journal of Energy, (2010), 13 (1)

work page 2010
[3]

Who’d have thought running an electricity grid and running a railway would have so much in common

A self -confessed transport geek, “Who’d have thought running an electricity grid and running a railway would have so much in common”, Transport Trends, (2012)

work page 2012
[4]

Hourly Electricity Load Forecasting: An Empirical Application to the Italian Railways

M. Centra , "Hourly Electricity Load Forecasting: An Empirical Application to the Italian Railways", World Academy of Science , Engineering and Technology, (2011), 56

work page 2011
[5]

P. K. Nag, Power Plant Engineering, Tata McGraw -Hill Education, (2008)

work page 2008
[6]

Molderink, Demand-Side Energy Management, University of Twente, Enschede, Netherlands

A. Molderink, Demand-Side Energy Management, University of Twente, Enschede, Netherlands

work page
[7]

The electrical load management, energy management and administrative strategic approach

F. Malekpour Farsadi, N. Eskandari and H. Siahi, "The electrical load management, energy management and administrative strategic approach", Second Conference Executive Management , (2011)

work page 2011

Show all 16 references

[8]

New Method for Energy - Efficient Train Operation on Commuter Rail Networks

H. Har amina, Mandić M. and Nikšić M., “New Method for Energy - Efficient Train Operation on Commuter Rail Networks”, TechnickiVjesnik, (2012), 19(4), 801-806

2012
[9]

Energy Efficiency on Train Control: design of Metro ATO driving and impact of energy accumulation devices

M. Domínguez, A. P. Cucala, A.Fernández, R.R. Pecharromán & J. Blanquer, “Energy Efficiency on Train Control: design of Metro ATO driving and impact of energy accumulation devices”, 9th World Congress on Railway Research, (2011)

2011
[10]

Recent Energy Saving Technologies on Railway Traction Systems

K. Kondo, “Recent Energy Saving Technologies on Railway Traction Systems”, IEEJ Transactions on Electrical and Electronic Engineering, (2010)

2010
[11]

Behbahani and A

M. Behbahani and A. Fatehi, Fundamental and Technology of Induction Motor Control, First edition, K. N. Toosi U. of Technology Publisher, (2011), ISBN: 978-964-8703-99-3

2011
[12]

Energy-saving Technology for Railway Traction Systems Using Onboard Storage Batteries

M. Shimada, Y. Miyaji, T. Kaneko & K. Suzuki, “Energy-saving Technology for Railway Traction Systems Using Onboard Storage Batteries”, Hitach review, (2012), 61 (7)

2012
[13]

New Technologies for Powering Public Transport Vehicles

A. Gheorghiur and M. Minea, “New Technologies for Powering Public Transport Vehicles”, Annals for the University of Croatia , Electrical Engineering series, (2010), 34

2010
[14]

Report of operation and activities of Consumption management in the year 2012

"Report of operation and activities of Consumption management in the year 2012", Tehran Regional Electric Company, (2012)

2012
[15]

Study of increasing Load Factor strategi es using demand management in over one megawatt plants in the Mazandaran

Hashemi Tehrani and Mohammadian Bisheh, "Study of increasing Load Factor strategi es using demand management in over one megawatt plants in the Mazandaran", 11th International Conference on Electricity Distribution Networks, Mazandaran, (2006)

2006

Pith tools

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

[1] [16]

High proportion of peak load losses in the total electrical energy losses

A. Sharifi Yazdi, “High proportion of peak load losses in the total electrical energy losses", Second National Energy Conference , (1999)

work page 1999

[2] [1]

Energy Balance Report 2010

Iranian Ministry of Energy, Department of Power and Energy, "Energy Balance Report 2010", (2010)

work page 2010

[3] [2]

Forecasting Electrical energy demand of the transport sector of Iran using MAED model by the year 2016

O. Shah Hoseini, Z. Soleimanian, R. Goodarzi and M. Alimardani, "Forecasting Electrical energy demand of the transport sector of Iran using MAED model by the year 2016", Journal of Energy, (2010), 13 (1)

work page 2010

[4] [3]

Who’d have thought running an electricity grid and running a railway would have so much in common

A self -confessed transport geek, “Who’d have thought running an electricity grid and running a railway would have so much in common”, Transport Trends, (2012)

work page 2012

[5] [4]

Hourly Electricity Load Forecasting: An Empirical Application to the Italian Railways

M. Centra , "Hourly Electricity Load Forecasting: An Empirical Application to the Italian Railways", World Academy of Science , Engineering and Technology, (2011), 56

work page 2011

[6] [5]

P. K. Nag, Power Plant Engineering, Tata McGraw -Hill Education, (2008)

work page 2008

[7] [6]

Molderink, Demand-Side Energy Management, University of Twente, Enschede, Netherlands

A. Molderink, Demand-Side Energy Management, University of Twente, Enschede, Netherlands

work page

[8] [7]

The electrical load management, energy management and administrative strategic approach

F. Malekpour Farsadi, N. Eskandari and H. Siahi, "The electrical load management, energy management and administrative strategic approach", Second Conference Executive Management , (2011)

work page 2011