{"id":"2fe0e8ba-c37a-494f-bb29-d262130401e1","arxiv_id":"1908.01111","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"As power system loads transition from constant power to constant impedance, equilibrium counts and the topological complexity of the algebraic manifold increase dramatically, but mostly on unstable components.","lead":"This paper uses numerical simulations to show that changing the load model in power systems from constant power to constant impedance changes the number of possible operating states and the shape of the system's mathematical solution surface. It finds that a mostly-impedance load creates more solutions and more complex geometric structures than a mostly-power load, but most of the added complexity appears on unstable parts of the system.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Topological claims ('sphere', 'glued tori', 'quotient space') are inferred from plotted 2D slices of the 9-bus manifold, with no homology computation or sampling protocol given, so the main discovery is not yet substantiated.","rationale":"The reader's weakest assumption identifies the same load-bearing concern: the topological mutation claims rest on visual inspection of 3D plots rather than on a rigorous computation of topological invariants. My reading strengthens this concern slightly by noting that the plotted surfaces are 2D renderings of a 3-dimensional algebraic manifold, and that the periodic angle coordinates make 'homeomorphic to a sphere' especially delicate without an explicit boundary-identification rule. However, the central claim of the paper is not solely topological: the equilibrium-count increases (Fig. 1), the non-monotonic solution-pair behavior (Fig. 2), and the stability-boundary transitions (Fig. 4) are numerical observations that can stand even if the topological language is softened. The abstract itself concedes that most changes occur on unstable components, which tempers practical significance but does not invalidate the mathematical observations. Therefore the appropriate response is not rejection but conditional acceptance: the authors should verify the homology of the algebraic manifold with a reproducible computation, or downgrade the topological statements to qualitative geometric descriptions. I agree with the reader's conditional verdict and do not propose moving it.","tokens_in":12416,"tokens_out":5098,"duration_ms":54735,"concrete_test":"Transform (7c)-(7e) for the 9-bus case into polynomial equations by substituting x=cos δ, y=sin δ (adding x^2+y^2=1), and for α ∈ {0, 0.2, 0.4, 0.6, 0.8, 0.931, 1.0} generate a dense real sample of the algebraic set in (δ21, δ31, V9) using the same continuation method as in the paper. Run persistent homology (e.g., Ripser) on the sampled point cloud and report Betti numbers β0, β1, β2 for the component containing the high-voltage SEP, following the same boundary-identification convention used in Fig. 5. If the Betti numbers do not match a sphere at α=0 and two glued tori at α=0.6, Section V's topological mutations are artifacts; the authors should either supply the exact sampling and identification protocol or soften the claims to qualitative geometry.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central 'major discovery' is that the algebraic manifold's topology mutates from a sphere to glued tori and eventually a quotient space as the load becomes impedance-dominated (Section V, Figs. 5-6). This is the paper's headline claim, but it is supported only by visual inspection of pink surfaces in the (δ21, δ31, V9) coordinate box. Two problems make the inference insecure. First, for the modified 9-bus system the algebraic constraints (7c)-(7e) have Ng=3 free dimensions after the COI reduction, so the full algebraic manifold is 3-dimensional; the figures render 2D surfaces, which must be either a section or a projection of that manifold. The paper never states which, nor how the surfaces were sampled, so the depicted object is not shown to be the full manifold whose topology is claimed. Second, the angle coordinates are periodic, yet the plots show one fundamental domain without specifying boundary identifications. A component that looks like a sphere in one box can become a torus once 2π replicas are identified, exactly the operation Fig. 5(f) invokes ('connects to its modulo 2π replica'). Therefore the statements 'homeomorphic to a sphere', 'two tori glued together', and 'quotient space of a few tori' are not inferable from the plots. The abstract's own caveat that changes occur mostly on unstable components limits practical implications, but the mathematical claim itself remains unverified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper investigates how the composition of static load models—constant power versus constant impedance, parametrized by α in Eq. (9)—affects the equilibria, stability regions, and algebraic manifolds of power system differential-algebraic equation (DAE) models. Using modified 5-, 9-, and 14-bus test systems in the center-of-inertia frame, the authors report that as the load becomes impedance-dominated, the number of equilibria increases drastically but non-monotonically, the stability region expands and its boundary shifts from singular surfaces to stable manifolds of type-1 UEPs, and the algebraic manifold is claimed to undergo topological mutations from a sphere to glued tori and eventually a quotient space, with most changes occurring on unstable components. The high-voltage solution is held fixed by construction through the impedance normalization in Eq. (10). The paper is primarily an exploratory numerical study with several counter-intuitive observations.","tokens_in":12702,"tokens_out":9445,"duration_ms":94178,"significance":"If substantiated, the reported topological mutations of the algebraic manifold would be a novel and counter-intuitive contribution to the power systems DAE literature, with potential implications for direct stability analysis and load-model selection in transient stability studies. The paper also has notable strengths: a systematic continuous parameter sweep over load-model composition, explicit acknowledgment that the high-voltage solution is fixed by construction, use of multiple test systems, and an honest limitation statement that most changes occur on unstable components. However, the central topological discovery is currently presented as a visual interpretation of sampled renderings; the evidence does not yet support the strength of the claims made in the abstract and Section V.","major_comments":[{"comment":"The central claim that the algebraic manifold's topology mutates from a sphere to glued tori and eventually a quotient space is not supported by the evidence presented. The figures show two-dimensional surfaces rendered in the (δ21, δ31, V9) box, but the algebraic constraints (7c)-(7e) live in a higher-dimensional space; the manuscript never states whether the plotted object is the full algebraic manifold, a section, or a projection, nor does it give the sampling procedure. Furthermore, the angle coordinates are periodic, and the plots show one fundamental domain without specifying boundary identifications; a component that appears as a sphere inside one box can become a torus once 2π replicas are identified (as Fig. 5(f) itself invokes). To make the topology claim load-bearing, the authors need either a precise mathematical definition of the plotted set plus computational verification of the claimed homeomorphism types (e.g., via triangulation and homology computation with certified bounds), or a reformulation of the claim as a visual observation about a specific slice.","section":"Section V, Figs. 5-6"},{"comment":"The quantitative claims that the number of equilibria increases drastically and non-monotonically with impedance proportion rest on the assumption that the method of [21,22] computes all equilibria. No completeness certificate is provided for the 5-, 9-, and 14-bus systems studied; if the method occasionally misses solutions or produces spurious ones, the reported non-monotone dips (e.g., Fig. 2(a) around 18.81% impedance) and the doubling of solutions in Fig. 1(c) could be numerical artifacts. The authors should provide evidence of completeness, for instance by comparing the counts against known total-degree bounds for the equivalent power flow equations, or by validating the method on a system where the full solution set can be independently verified.","section":"Section III, Figs. 1-2"},{"comment":"The stability-region identification in Algorithm 1 traces each of M rays from the high-voltage SEP in the (δ21, δ31) plane and assumes that the intersection of the stability boundary with each ray is a single point (equivalently, that the stability region is star-shaped in this projection). This assumption is not justified and is generally false for power system stability regions, which can have complicated non-convex shapes. The claims that the stability region 'expands' and that its boundary transitions from the singularity boundary to the stable manifolds of type-1 UEPs are therefore only established for this ray-slicing approximation. Please either justify the star-shaped assumption for these cases or report the dependence of the conclusions on the choice of M and on the particular projection used.","section":"Section IV, Algorithm 1 and Fig. 4"}],"minor_comments":[{"comment":"In Eq. (8), the reactive current coefficient is written as 'Id,j' in (8b) but the text says 'Iq,j = 0'; the symbol should be corrected to 'Id,j'. Also, the phrase 'to acquire a uni-directional change of load model' should read 'to achieve a uni-directional change of the load model'.","section":"Section II-B, Eq. (8)"},{"comment":"The sentence 'neither a type-1 equilibrium point (on the stable component of the algebraic manifold) should admit one low bus voltage, nor a solution with only one low bus voltage should be type-1' is grammatically confusing and reads as a normative claim. Please state explicitly that this is an empirical observation from the tested cases and not a general theorem.","section":"Section III-B"},{"comment":"The term 'stable component of the algebraic manifold' is used repeatedly but never formally defined. Please provide a definition (e.g., the connected component containing the high-voltage stable equilibrium) and clarify whether the classification of components as stable/unstable is independent of the load-model parameter.","section":"Sections III-B and V"},{"comment":"The captions of Figs. 5 and 6 should state whether the plotted pink surface is a slice of the algebraic manifold or a projection, which variables are held fixed, and how the surface was sampled (grid resolution, solver used, treatment of multiple solution branches).","section":"Section V, Figs. 5-6 captions"},{"comment":"There are several typos that should be corrected, including 'Appearently' (Section II-A), 'formualted' (Section II), 'equilibira' (Section III-A), and 'Eqt.' instead of 'Eq.'.","section":"Throughout"},{"comment":"The normalization in Eq. (10) fixes the high-voltage solution by construction. Since several qualitative findings are drawn from sweeping α, a brief sensitivity study with respect to the base operating point, or an explicit scoping statement that results apply only under this normalization, would help the reader judge the generality of the conclusions.","section":"Section II-B, Eq. (10)"}],"recommendation":"major_revision","confidential_remarks":"For the editor: this is an exploratory numerical study whose headline claim—topological mutation of the algebraic manifold—is not yet supported to the standard needed for a rigorous journal publication. The authors build directly on their own previous method for enumerating power-flow solutions [21,22], and the completeness of that method is central to the equilibrium-count findings. If the authors can provide computational topology verification or downgrade the claim to a description of rendered slices, the paper could become publishable. The paper also does not discuss the fit between its empirical approach and the journal's expectations for theoretical rigor."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the equilibrium and stability-region observations in this paper are worth a look, but the headline topology claims ('sphere mutates to glued tori, then quotient space') are not supported by the evidence presented. I would not accept that part as a result; I'd treat it as a suggestive picture.\n\nWhat's actually new: the paper takes the single-machine study in [15] and systematically sweeps the power/impedance mix in 5-, 9-, and 14-bus systems. The non-monotone change in the number of equilibria (Figures 1-2) is a concrete, checkable observation. The finding that type-1 UEPs on the stable component don't line up with single low-voltage buses (Table I) is also useful for people who use Newton's method with heuristic initial guesses. And the stability-region transition from singularity-boundary-dominated to UEP-stable-manifold-dominated (Figure 4) is a clean numerical demonstration. The authors are also honest about the big caveat: most of the complexity appears on unstable components, so practical dynamic behavior on the stable component is broadly unchanged.\n\nWhere it's soft: Section V. The algebraic manifold is defined by the DAE constraints in the full variable space, but Figures 5-6 plot a surface in the (δ21, δ31, V9) slice. The paper never says how that surface was sampled, whether it's a section or projection, or how boundary identifications at 2π are handled. Without that, 'homeomorphic to a sphere', 'two tori glued together', and 'quotient space' are inferences from pictures, not mathematical statements. A homology computation or at least a defined sampling grid with identifications would fix this; alternatively, the language should be downgraded to 'the projected surface appears to develop tunnels'. The enumeration of equilibria also depends on the solver in [21], [22] — no code or data shipped, so completeness is an act of trust. Minor issue: the 'stability region' in Section IV is a slice on the zero-speed plane, which they acknowledge, so it's fine, but keep that in mind.\n\nNet: it's a numerical study with a solid empirical core and one overreach. For a power-systems audience, the equilibrium and boundary observations are worth engaging. The topological claims need either real verification or explicit softening before they should propagate. If I were editor, I'd send it to review rather than desk-reject, with the expectation that the authors either compute homology or rewrite Section V as qualitative.","headline":"Equilibrium and stability-region observations are useful numerics; the algebraic-manifold topology claims are pictures, not results.","tokens_in":13193,"tokens_out":3283,"would_cite":false,"duration_ms":32068,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that shifting loads from constant power to constant impedance sharply increases the number of equilibria and changes the topology of the algebraic manifold, while leaving the stable component mostly unchanged.","keywords":["load model","differential-algebraic equations","equilibrium","stability region","algebraic manifold","constant power load","constant impedance load","topology"],"falsifier":"Compute the real algebraic variety of the 9-bus DAE constraints at 40%, 60%, and 93.1% impedance using certified real-root isolation or a cylindrical algebraic decomposition, and compare its Euler characteristic or homology to the claimed sphere, two glued tori, and quotient-space descriptions. A discrepancy would show the reported topological mutation is a plotting or sampling artifact.","tokens_in":12230,"feed_emoji":"⚡","tokens_out":4426,"duration_ms":41142,"temperature":0.7,"pith_summary":"This paper asks how the choice of load model changes the global structure of a power system's differential-algebraic equations, and it reports that the answer is much more dramatic than the load-model literature suggests. As loads are continuously morphed from constant power to constant impedance, the number of equilibrium solutions rises sharply (in one 14-bus test, doubling between 80% and 90% impedance), the singular surface shrinks and eventually vanishes, and the stability region grows until it is bounded by stable manifolds of type-1 UEPs instead of by singularities. The paper's headline discovery is that the algebraic manifold itself changes topology, from a sphere-like object to two glued tori and then to a more complicated quotient space, with the most intricate structure appearing when impedance dominates and constant power is nearly gone. Crucially, the paper argues that nearly all of this added complexity lives on unstable components or near singular boundaries, while the stable component keeps a single stable equilibrium point in every tested case. If correct, this means that practical transient-stability conclusions based on a chosen load model may be robust in the stable region even though the underlying manifold is wildly changing elsewhere.","feed_headline":"Impedance-heavy loads multiply power-grid equilibria","feed_subtitle":"A 9-bus study finds the algebraic manifold morphs from sphere to glued tori as constant power fades.","key_machinery":"The central object is the algebraic manifold of the DAE: the solution set of the algebraic constraints (voltage-magnitude equations at generator buses and real/reactive power balances at load buses) in the space of relative generator angles and bus voltages. The mechanism that drives the study is a linear interpolation parameter α, where the load at each bus is α times constant power plus (1−α) times equivalent constant impedance, so α=1 is pure constant power and α=0 is pure impedance. Equilibria are found by solving an equivalent power-flow problem in the COI frame, with each equilibrium typed by the eigenvalues of the reduced Jacobian; stability-region boundaries are identified numerically by a ray-search algorithm that distinguishes singular-surface crossings from crossings of stable manifolds of type-1 UEPs. The load-bearing computational tools are the multi-solution power-flow solver of [21] and [22] and the DAE stability-boundary theory of [24] and [25].","core_discovery":"Using the IEEE 9-bus (and 5- and 14-bus) systems in a center-of-inertia classical DAE model, the paper claims that continuously replacing constant-power loads with equivalent constant-impedance loads produces a sequence of structural changes in the solution set. The number of power-flow-like equilibria increases by a large factor but not monotonically; equilibria of different types appear and collide, and pairs of real solutions vanish into the complex plane and later reappear. On the stable component of the algebraic manifold there remains exactly one stable equilibrium point (the high-voltage solution), and the number of equilibria there stays small. The stability boundary, initially the singular surface of the DAE, expands and is progressively taken over by the stable manifolds of type-1 unstable equilibria until, at pure impedance, the singular surface disappears. The algebraic manifold, plotted in the angle-angle-voltage slice, enlarges and connects to its modulo-2π replicas, and its topology is reported to mutate from sphere to glued tori to a quotient space; the authors emphasize this is counterintuitive because constant impedance is usually considered the simpler load model.","pith_inferences":["If the topological mutation is real, the algebraic manifold of impedance-loaded systems is not simply connected; direct methods that rely on the stable component's basin structure may need to account for trajectories that wind around tunnels before converging.","The non-monotone equilibria count suggests that homotopy-continuation or holomorphic-embedding solvers should expect saddle-node-style collisions of real solution pairs as load composition changes, providing a testable prediction for other bus systems.","One could test whether the same complexity increase appears under ZIP loads with nonzero current terms; the paper deliberately omits constant-current components, so its conclusions may be a best-case for topology change.","Practical load-model identification may benefit: if stable-component behavior is robust while unstable components proliferate, then matching measured transient behavior may not require exactly reproducing the full algebraic manifold."],"forward_implications":["The number of equilibrium solutions a power-flow-style solver must contend with grows sharply as impedance dominates, so contingency screening based on a single load model could miss many operating points.","Stability regions under constant-power loads are bounded by the singularity surface; under pure impedance they are bounded by stable manifolds of type-1 UEPs, so the governing mechanism of transient stability changes with load composition.","A low voltage at a single bus is not a reliable signature of a type-1 UEP on the stable component, and type-1 UEPs can show low voltages at several buses simultaneously.","The algebraic manifold connects to its modulo-2π replicas under impedance-dominated loads, implying the DAE dynamics can pass through multiple 2π angular sweeps and still settle at an equilibrium.","Because the added equilibria and topological changes sit on unstable components, stable-component behavior (and the uniqueness of the high-voltage SEP) may remain largely unaffected, though this is observed, not proven."],"supporting_citations":[{"why":"Earlier DAE load-model study on a single-machine infinite-bus system; supplies the singular-surface stability-boundary picture that this paper extends.","marker":"[15]"},{"why":"Holomorphic-embedding continuation method used to enumerate all equilibria for each load model.","marker":"[21]"},{"why":"Efficient method to locate all load-flow solutions, the basis for finding multiple equilibria in this paper.","marker":"[22]"},{"why":"DAE stability-region theory identifying singular surfaces as part of stability boundaries.","marker":"[24]"},{"why":"Taxonomy of constrained nonlinear system dynamics, used to classify stability-boundary components.","marker":"[25]"},{"why":"Comparison of generator and load models on transient-stability results, cited for the expected stability-region expansion.","marker":"[27]"}],"fun_headline_variants":["Impedance-heavy loads multiply grid equilibria, alter manifolds","Load model shift morphs grid topology from sphere to tori","Constant impedance spawns extra equilibria, surprises topology","Impedance loads: complex topology, many equilibria"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The topological claims (sphere, glued tori, quotient space) rest on visual inspection of 3D plots of a single angle-angle-voltage slice, sampled numerically, without a formal proof that these plots correctly capture the global topology of the full algebraic variety.","fun_headline_variants_meta":{"raw":{"variants":["Impedance-heavy loads multiply grid equilibria, alter manifolds","Load model shift morphs grid topology from sphere to tori","Constant impedance spawns extra equilibria, surprises topology","Impedance loads: complex topology, many equilibria"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000687,"raw_usage":{"total_tokens":3134,"prompt_tokens":987,"completion_tokens":2147,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":603,"completion_tokens_details":{"reasoning_tokens":2078}},"tokens_in":603,"tokens_out":2147,"duration_ms":17031,"temperature":1.0,"reasoning_tokens":2078,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:23:09.873587+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the real algebraic variety of the 9-bus DAE constraints at 40%, 60%, and 93.1% impedance using certified real-root isolation or a cylindrical algebraic decomposition, and compare its Euler characteristic or homology to the claimed sphere, two glued tori, and quotient-space descriptions. A discrepancy would show the reported topological mutation is a plotting or sampling artifact.","supporting_citations":[{"cited_title":"Geometrical structure of constraint manifold in power system differential-algebra ic model,","cited_arxiv_id":null,"evidence_quote":"Earlier DAE load-model study on a single-machine infinite-bus system; supplies the singular-surface stability-boundary picture that this paper extends."},{"cited_title":"Holomorphic embedding based continu ation method for identifying multiple power flow solutions,","cited_arxiv_id":null,"evidence_quote":"Holomorphic-embedding continuation method used to enumerate all equilibria for each load model."},{"cited_title":"An efficient method to locate all the load flow solutions-revisited,","cited_arxiv_id":null,"evidence_quote":"Efficient method to locate all load-flow solutions, the basis for finding multiple equilibria in this paper."},{"cited_title":"Stability regions for differential-algebraic systems,","cited_arxiv_id":null,"evidence_quote":"DAE stability-region theory identifying singular surfaces as part of stability boundaries."},{"cited_title":"Dynamics of large constrained nonlinear systems-a taxonomy theory [po wer system stability],","cited_arxiv_id":null,"evidence_quote":"Taxonomy of constrained nonlinear system dynamics, used to classify stability-boundary components."},{"cited_title":"Effect of gene rator models and load models on the results of the transient stability ana lysis of a power system,","cited_arxiv_id":null,"evidence_quote":"Comparison of generator and load models on transient-stability results, cited for the expected stability-region expansion."}],"review_version":1}