{"id":"dada09e8-41fd-4ebb-8afa-c8d700b106e6","arxiv_id":"2505.23435","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"On the IEEE European LV test feeder, the line-voltage-based LVUR tracks the true VUF closely, the phase-voltage-based PVUR1 and PVUR2 deviate strongly, and PV integration changes both unbalance and metric accuracy depending on placement.","lead":"This paper compares five standard ways of measuring voltage unbalance on a realistic low-voltage distribution grid with and without rooftop solar. It finds that the metric choice matters, and that solar can improve or worsen unbalance depending on where it is connected.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Scenario III in Table V contradicts the claim that all indices improve when VUF falls: mean VUF decreases by 0.074 while mean |PVUR2−VUF| rises by 0.597.","rationale":"The reader's conditional verdict and representativeness concern are reasonable: the paper samples 9 of 55 buses and gives no model/code. However, the most load-bearing problem lies closer to the argument itself. The paper's Section IV/V conclusion asserts a monotonic coupling between VUF direction and accuracy direction for all indices, and its own Tables III–V provide a counterexample in Scenario III. This is a direct arithmetic inconsistency, independent of whether the selected buses or scenarios are representative. The broader qualitative claims (LVUR approximates VUF better than PVUR1/PVUR2 because line voltages encode phase-angle differences; CIGRE equals VUF exactly) are well-supported by prior literature and by the structure of the definitions, so a full rejection would be disproportionate. A conditional verdict remains appropriate, but the requested revisions should include correcting or qualifying the all-indices-improve claim and reporting the distribution across buses rather than only the selected 9-bus summary. My disagreement with the reader's weakest_assumption is therefore not about the validity of the sampling concern, but about which issue is decisive: the reported numbers themselves already fail to support one of the central conclusions.","tokens_in":6777,"tokens_out":6553,"duration_ms":61854,"concrete_test":"Recompute Table V directly from Tables III and IV as after-minus-before mean values for each scenario. The decisive arithmetic check is Scenario III: mean VUF change = 1.484 − 1.558 = −0.074, and mean |PVUR2−VUF| change = 8.350 − 7.753 = +0.597. If these entries are correct, the claim 'when PV integration reduces VUF, all indices improve' is false as stated. To settle the issue, the authors should either report per-bus sign patterns for all 9 buses (for example, count buses where VUF decreases but PVUR2 error increases) or revise the conclusion to state that index-specific accuracy can diverge from the direction of VUF change.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's central empirical dependency, stated in Section IV and the Conclusion, is that when PV integration reduces VUF, the accuracy of all evaluated indices improves, and that when VUF increases, accuracy deteriorates. The paper's own reported numbers contradict the first direction. In Scenario III, Table III gives a mean VUF of 1.558% before PV and 1.484% after PV, an improvement of 0.074 percentage points. Table IV gives mean |PVUR2−VUF| of 7.753% before and 8.350% after, a deterioration of 0.597 percentage points, and the maximum error also rises from 10.125% to 11.629%. Thus a phase-voltage-based index becomes less accurate while the true unbalance factor improves, which is exactly the opposite of the stated dependency. The text dismisses Scenario III as a 'minimal change,' but no threshold is provided and the VUF movement is not zero. The conclusion is therefore unsupported by the reported data as written; this is an internal inconsistency, not merely an external generalizability question.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper compares five voltage unbalance metrics (VUF, LVUR, CIGRE, PVUR1, and PVUR2) under three load scenarios in the IEEE European low-voltage test feeder, then studies how the addition of 40 single-phase PV units affects both the level of true unbalance (measured by VUF) and the accuracy of the approximate metrics. It reports numerical bounds of each metric relative to VUF for VUF ranges of 1–2% and 2–3%, and concludes that LVUR is a reasonable approximation of VUF, that CIGRE is an exact reformulation, that PVUR1 and PVUR2 are less reliable because they neglect phase-angle deviations, and that PV integration improves index accuracy whenever it reduces VUF. The analysis is carried out in OpenDSS, with results reported for nine selected buses out of the feeder's 55 load buses.","tokens_in":7029,"tokens_out":4526,"duration_ms":44031,"significance":"The topic is practically relevant: distribution engineers need guidance on when inexpensive magnitude-only unbalance metrics can substitute for the sequence-based VUF in networks with high DER penetration. The simulation setup is appropriate, the use of absolute error relative to VUF is a reasonable evaluation approach, and the qualitative finding that LVUR is close to VUF in this feeder is credible from the reported mean errors (≤0.1 percentage points in Table IV). The paper also correctly leverages the known exactness of the CIGRE metric. However, the paper's additional claim of a deterministic monotonic relation between VUF changes and index-accuracy changes is contradicted by its own data, and the generality of the numerical results is limited by the nine-bus selection and the hand-picked load/PV scenarios. If the analysis were corrected and its scope stated honestly, the paper would be a useful case-study contribution, though it does not introduce a new theoretical result.","major_comments":[{"comment":"The central empirical claim, stated in Section IV and the Conclusion, that \"when PV integration reduces VUF, all indices improve\" is directly contradicted by Scenario III. Table III shows the mean VUF decreases from 1.558% to 1.484% (an improvement of 0.074 percentage points), while Table IV shows the mean |PVUR2−VUF| error increases from 7.753% to 8.350% and the maximum error rises from 10.125% to 11.629%. Table V reports a deterioration of +0.597 for PVUR2 in Scenario III. The text describes Scenario III as a \"minimal change,\" but the VUF movement is nonzero and the PVUR2 error change is larger than in Scenario I (+0.277) and comparable in sign to the deterioration in Scenario I. This internal inconsistency means the stated dependency is not supported by the reported data; the authors must either revise the claim to acknowledge that PVUR2 can worsen while VUF improves, or add a condition (e.g., a threshold on phase-angle deviation changes) that explains the discrepancy.","section":"Section IV, Tables III and V"},{"comment":"The numerical bounds in Tables I and II are load-bearing for the paper's conclusion that LVUR is a reliable approximation (upper bound near 1.8% for the 2–3% VUF range), yet the method used to compute them is not described. The text gives only the input ranges in Eqs. (17)–(18) and the inequalities (19)–(21). It is unclear whether the bounds are exact extrema over the continuous domain, results of a Monte Carlo/random search, or solutions of a formal optimization problem. No step sizes, sampling density, optimizer, or convergence criteria are reported. Without this information readers cannot reproduce or verify the values 0.866, 1.005, 10.728, 16.092, etc., nor assess whether the LVUR upper bound is a true worst case or an artifact of an incomplete search. Please specify the computational method and, if the bounds are not proven, state explicitly that they are empirical estimates rather than exact limits.","section":"Section II-F, Tables I and II"},{"comment":"The quantitative conclusions about feeder-wide unbalance and DER effects are based on nine selected buses (beginning, middle, and end of each of three zones) out of 55 load buses, and on three hand-picked load scenarios plus a single PV phase-allocation pattern. The paper asserts that these buses \"represent the overall situation of each zone\" without supporting evidence. Since the reported means in Tables III–V are averaged over only nine buses, a different choice of representative buses could change the results—for instance, whether Scenario II's mean VUF drop from 1.255% to 0.702% is typical of the whole feeder. The authors should either justify the selection, report statistics for all 55 buses, or perform a sensitivity analysis over bus choices. Likewise, the three load scenarios and the single PV allocation are not varied systematically, so the paper should acknowledge that the DER-related conclusions are case-study specific rather than general.","section":"Section III, Fig. 2, and Section IV"}],"minor_comments":[{"comment":"The first row of Table V is labeled \"LVUF\" but should be \"LVUR\" (Line Voltage Unbalance Ratio); the same typo appears in the surrounding text in Section IV.","section":"Table V"},{"comment":"The typesetting of Eq. (8) is inconsistent: \"V 4 ab\" and \"V 2ca\" should be V^4_ab and V^2_ca, and the spacing in the denominator should be fixed for readability.","section":"Equation (8)"},{"comment":"The notation \"VU Ind.\" is undefined; introduce an index variable, e.g., I ∈ {LVUR, CIGRE, PVUR1, PVUR2}, and rewrite Eqs. (19)–(21) in terms of that variable to avoid ambiguity.","section":"Section II-F, Eqs. (19)–(21)"},{"comment":"The figures showing index values across buses are hard to read in print: markers are not labeled with bus numbers, and the legends do not identify which curves correspond to which zone. Please add bus numbers or provide a supplementary table of values.","section":"Figures 3–8"},{"comment":"The conclusion repeats the unsupported monotonic claim described in Major Comment 1; it should be rewritten to state what the data actually show, including the Scenario III exception for PVUR2.","section":"Section IV, Conclusion"},{"comment":"The OpenDSS model and the load/PV scenario data are not made available, which limits reproducibility; consider providing a data and code link in the final version.","section":"Reproducibility"}],"recommendation":"major_revision","confidential_remarks":"The paper fits eess.SY and addresses a practical problem, but the internal inconsistency in the DER-accuracy claim is significant and must be fixed before publication. I recommend the editor require the authors to also provide the OpenDSS model and raw results as supplementary material, and to either prove or explicitly label the bounds in Tables I–II as empirical estimates."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a confirmatory OpenDSS case study on standard voltage unbalance metrics. The qualitative ranking it reports—LVUR is a decent approximation, CIGRE is exact, PVUR1/PVUR2 are less reliable—is almost certainly correct. The problem is the paper's central DER conclusion is contradicted by its own numbers in Scenario III.\n\nWhat is actually new: not much conceptually. The definitions and bounds come from earlier work [12]–[14], and CIGRE equals VUF is a known identity. The contribution is a worked simulation on the IEEE European LV feeder with three load scenarios and one PV allocation pattern. That is a legitimate extension, and the figures are consistent with prior expectations. The bounds tables match earlier results, but the search over voltage magnitudes and phase deviations is not described, so I cannot verify the numbers.\n\nSoft spots, in order of severity:\n1. Scenario III inconsistency. Table III shows mean VUF moving from 1.558% to 1.484% after PV, an improvement. Table IV shows mean |PVUR2−VUF| moving from 7.753% to 8.350% and the maximum from 10.125% to 11.629%, a clear deterioration. The conclusion's blanket claim that when VUF improves all indices improve is false in the paper's own data. Calling Scenario III a 'minimal change' does not fix it; 0.074 points is still an improvement, and the error moved 0.597 points in the opposite direction.\n2. No code or data. The OpenDSS model is not shipped, so the 9-of-55 bus selection and the PV phase allocation cannot be audited.\n3. Generalization. Three hand-picked scenarios and nine buses are too thin to support universal statements about DER effects.\n\nWho is this for: engineers wanting a worked example of how these metrics behave on a standard LV feeder. Not for readers looking for new theory. I would send it to peer review rather than desk reject, because the flaw is fixable and the case study has value—but the authors need to resolve the Scenario III contradiction and document the search method and model before the central conclusion can be trusted.","headline":"Confirmatory OpenDSS case study on voltage unbalance metrics whose central DER conclusion is contradicted by its own Scenario III numbers, though the metric ranking itself is likely correct.","tokens_in":7536,"tokens_out":5338,"would_cite":false,"duration_ms":49220,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper shows that on a simulated European low-voltage distribution feeder, the line-voltage unbalance ratio (LVUR) is a reasonable approximation of the true sequence-based voltage unbalance factor, while the two phase-voltage metrics…","keywords":["voltage unbalance","voltage unbalance factor","LVUR","PVUR","distributed energy resources","photovoltaic integration","power quality","low-voltage distribution network"],"falsifier":"Measure VUF, LVUR, CIGRE, PVUR1, and PVUR2 at all 55 buses of the European low-voltage feeder across load allocations drawn from a realistic distribution and across several random PV phase placements; the paper's ranking would be falsified if any phase-voltage-based metric tracks VUF more closely than LVUR on average, or if the direction of the PV effect reverses for the same scenario when measured at non-selected buses.","tokens_in":6589,"feed_emoji":"⚡","tokens_out":10384,"duration_ms":96499,"temperature":0.7,"pith_summary":"The paper sets out to determine which of the common voltage-unbalance formulas engineers should trust in distribution networks with high solar penetration. Using the European low-voltage test feeder, it compares the sequence-based voltage unbalance factor (VUF) against the line-voltage ratio LVUR, the CIGRE factor, and two phase-voltage ratios, before and after adding 40 single-phase rooftop PV units. It finds that LVUR is a reasonable approximation of VUF, CIGRE is an exact reformulation, and the two phase-voltage metrics (PVUR1, PVUR2) can deviate from VUF by several percentage points because they ignore phase-angle deviations. It also finds that PV integration can raise, lower, or roughly leave unchanged the unbalance and the metrics' accuracy depending on how loads and PV are allocated across phases.","feed_headline":"Line-voltage metric tracks true unbalance; phase metrics miss by 11%","feed_subtitle":"On a simulated 416-V feeder, PV cuts unbalance in one load pattern, raises it in another.","key_machinery":"The central object is the voltage unbalance factor (VUF), computed as the ratio of negative-sequence to positive-sequence voltage magnitude, which the paper treats as the true definition. Around it, the paper evaluates LVUR (maximum line-voltage deviation over average line voltage), the CIGRE factor (an exact reformulation of VUF from line-voltage magnitudes), PVUR1 (maximum phase-voltage deviation over average), and PVUR2 (max-minus-min phase voltage over average). The comparison mechanism is the absolute error of each approximate index relative to VUF, computed for nine representative buses (beginning, middle, and end of three feeder zones) across three hand-specified unbalanced load scenarios, before and after adding 40 single-phase 2.5-kW grid-following PV units in power-flow simulations.","core_discovery":"The paper's central claim is that in the simulated European low-voltage network, LVUR is a reasonable approximation of the true definition (VUF), while PVUR1 and PVUR2 are less reliable because they exclude phase angle deviations; the CIGRE index is an exact reformulation of VUF and therefore matches it. The paper further claims that the effect of PV integration on both the level of unbalance and the accuracy of the indices is scenario-dependent: in Scenario I the average VUF rose and index accuracy fell, in Scenario II the average VUF fell and accuracy rose, and in Scenario III the change was negligible. These conclusions come from comparing each index's absolute error against VUF at nine selected buses over three load-distribution scenarios, using the European low-voltage test feeder simulated in a power-flow tool.","pith_inferences":["A natural next step would be to repeat the comparison at all 55 buses and under hundreds of random load and PV phase allocations; the paper's point estimates would then become distributional claims.","Because line-voltage metrics cannot see zero-sequence components, networks with grounded-wye or heavily single-phase loads might find LVUR too optimistic; a zero-sequence or neutral-current check would complement it.","The scenario-dependence means a single penetration label is not a reliable predictor of unbalance; phase-resolved allocation should be treated as a design variable in planning studies.","Operators without angle measurements could adopt the CIGRE factor instead of LVUR, since it is exact while still using only line-voltage magnitudes."],"forward_implications":["LVUR can serve as a reasonable proxy for the sequence-based VUF in low-voltage feeders when phase-angle measurements are unavailable, though it may underestimate VUF.","PVUR1 and PVUR2 are unreliable in feeders with significant phase-angle deviations; in the simulated scenarios their mean absolute error reached up to about 8.4 percentage points, and the largest single-bus error about 11.6 points.","The CIGRE factor reproduces VUF exactly without needing phase-angle measurements, making it the most accurate magnitude-only metric.","Integrating single-phase PV can statistically raise, lower, or leave unchanged the feeder's VUF depending on the phase allocation of loads and PV, so DER placement is a power-quality lever.","When PV integration reduces VUF, all approximate indices become more accurate; when it raises VUF, they become less accurate."],"supporting_citations":[{"why":"Defines the voltage unbalance factor (VUF) from symmetrical components, used as the true baseline in the comparison.","marker":"[5]"},{"why":"Defines the line voltage unbalance ratio (LVUR), one of the approximate metrics evaluated.","marker":"[7]"},{"why":"Provides the CIGRE unbalance factor, which the paper treats as an exact reformulation of VUF.","marker":"[8]"},{"why":"Defines the phase voltage unbalance ratio (PVUR1) evaluated in the comparison.","marker":"[9]"},{"why":"Defines the phase voltage unbalance ratio (PVUR2) evaluated in the comparison.","marker":"[10]"},{"why":"Supplies the observation that line-voltage measurements cannot capture the zero-sequence component, a known bias for LVUR.","marker":"[13]"},{"why":"Establishes the theoretical relationships and relative bounds among voltage unbalance definitions that the simulation results align with.","marker":"[14]"},{"why":"Provides the low-voltage test feeder network used for the power-flow simulations.","marker":"[15]"},{"why":"The power-flow simulation tool used to compute bus voltages in the test feeder.","marker":"[21]"}],"fun_headline_variants":["LVUR tracks true unbalance; phase metrics miss by 11%","PV sways unbalance both ways and fools the wrong indices","Which unbalance index is reliable? Depends on DER siting","CIGRE index equals VUF; PVUR variants fall short on phase","High-PV grids: choose LVUR, not phase metrics, for unbalance"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the nine selected buses, three load scenarios, and one PV phase-placement pattern faithfully represent the unbalance behaviour of the whole feeder and of real low-voltage networks; if those selections are not representative, the reported mean errors and DER effects will not generalize.","fun_headline_variants_meta":{"raw":{"variants":["LVUR tracks true unbalance; phase metrics miss by 11%","PV sways unbalance both ways and fools the wrong indices","Which unbalance index is reliable? Depends on DER siting","CIGRE index equals VUF; PVUR variants fall short on phase","High-PV grids: choose LVUR, not phase metrics, for unbalance"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000665,"raw_usage":{"total_tokens":2971,"prompt_tokens":817,"completion_tokens":2154,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":433,"completion_tokens_details":{"reasoning_tokens":2058}},"tokens_in":433,"tokens_out":2154,"duration_ms":19484,"temperature":1.0,"reasoning_tokens":2058,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T12:45:35.019549+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure VUF, LVUR, CIGRE, PVUR1, and PVUR2 at all 55 buses of the European low-voltage feeder across load allocations drawn from a realistic distribution and across several random PV phase placements; the paper's ranking would be falsified if any phase-voltage-based metric tracks VUF more closely than LVUR on average, or if the direction of the PV effect reverses for the same scenario when measured at non-selected buses.","supporting_citations":[{"cited_title":"National Electrical Manufacturers Association (NEMA), 1993","cited_arxiv_id":null,"evidence_quote":"Defines the line voltage unbalance ratio (LVUR), one of the approximate metrics evaluated."},{"cited_title":"IEC 61000-2-2, EMC Part 2-2: Environment Compatibility Levels for Low Frequency Conducted Disturbances and Signalling in Public Low-V oltage Power Supply Systems,","cited_arxiv_id":null,"evidence_quote":"Defines the voltage unbalance factor (VUF) from symmetrical components, used as the true baseline in the comparison."},{"cited_title":"A new simple and effective approximate formulation for the determination of three-phase unbalances by voltmeter method,","cited_arxiv_id":null,"evidence_quote":"Provides the CIGRE unbalance factor, which the paper treats as an exact reformulation of VUF."},{"cited_title":"IEEE, 1994","cited_arxiv_id":null,"evidence_quote":"Defines the phase voltage unbalance ratio (PVUR1) evaluated in the comparison."},{"cited_title":"IEEE Standard Test Procedure for Polyphase Induction Motors and Generators,","cited_arxiv_id":null,"evidence_quote":"Defines the phase voltage unbalance ratio (PVUR2) evaluated in the comparison."},{"cited_title":"Examination of the definitions of voltage unbalance,","cited_arxiv_id":null,"evidence_quote":"Supplies the observation that line-voltage measurements cannot capture the zero-sequence component, a known bias for LVUR."},{"cited_title":"On the relationships among different voltage unbalance definitions,","cited_arxiv_id":null,"evidence_quote":"Establishes the theoretical relationships and relative bounds among voltage unbalance definitions that the simulation results align with."},{"cited_title":"IEEE European Low V oltage Test Feeder,","cited_arxiv_id":null,"evidence_quote":"Provides the low-voltage test feeder network used for the power-flow simulations."},{"cited_title":"Introduction to OpenDSS","cited_arxiv_id":null,"evidence_quote":"The power-flow simulation tool used to compute bus voltages in the test feeder."}],"review_version":1}