EL26-67-000 · 2026-07-08 · Individual No Affiliation

Supplemental Technical Comment of L.M. Marlowe on the integration of large, high-density electric loads and the associated reliability, observability, and cost-allocation methodologies under consideration by the Commission

UNITED STATES OF AMERICA 
BEFORE THE 
FEDERAL ENERGY REGULATORY COMMISSION 
 
Integration of Large Loads and Associated Reliability and Cost-
Allocation Methodologies; 
Show-Cause Proceedings and Related Dockets 
) Docket No. EL26-67-000 
) Docket No. EL26-68-000 
) Docket No. EL26-69-000 
) Docket No. EL26-70-000 
) Docket No. EL26-71-000 
) Docket No. EL26-72-000 
) Docket No. RM26-4-000 
) Docket No. RP26-358-000 
 
SUPPLEMENTAL TECHNICAL COMMENT AND RESERVATION OF RIGHTS OF L.M. 
MARLOWE 
Date: July 8, 2026 
This submission is made in good faith by a private citizen and independent researcher, supported by extensive 
operational documentation and prior art. It is offered for the public benefit — to improve grid reliability, 
transparency in large-load cost allocation, and the protection of ratepayers from systemic waste and undue cost-
shifting. It is based on publicly available records, Commenter’s own timestamped technical documentation and 
prior-art filings, and technical analysis. 
To the Honorable Secretary, Federal Energy Regulatory Commission, 888 First Street, NE, Washington, 
DC 20426: 
Lisa Melton, filing as L.M. Marlowe (“Commenter”), respectfully submits this supplemental technical 
comment in the above-captioned proceedings concerning the integration of large, high-density electric 
loads and the associated reliability, observability, and cost-allocation methodologies under consideration 
by the Commission. Commenter is a party whose Motion to Intervene in Docket No. EL26-67-000, et al. 
was accepted for filing (Accession No. 202607025022, July 2, 2026). This comment supplements and 
elaborates upon Commenter’s prior submissions, and is offered for the Commission’s consideration in the 
captioned proceedings. Nothing herein is a legal conclusion; the technical characterizations are drawn 
from Commenter’s contemporaneous, timestamped record and document sequence, not causation. 
I. Interest of the Commenter 
Commenter is an independent author and developer of a published, timestamped analytical framework 
addressing deterministic reliability invariants, real-time observability of computational load behavior, 
dependency mapping, and the allocation of reliability and capacity burdens created by large 
computational loads. Commenter’s published work — including the Architecture of Dependency and 
Autonomy™, the MARLOWE™ Invariant-Based Diagnostic Framework, Intent-Aware Telemetry™, 
Continuous Invariant Validation™, Execution Sovereignty™, and the associated Ghost Load™ 
dependency analysis — bears directly on the reliability and cost-causation questions presented in these 
dockets. Commenter therefore has a direct and substantial interest in their outcome. Commenter states

that interest plainly: Commenter is the author and rights-holder of this framework and submits it for the 
Commission’s consideration in the belief that it responds to the reliability, observability, and cost-
causation questions these proceedings present. Commenter’s object in these proceedings is that the 
reliability of the system be protected and that the costs of large computational load be borne by those who 
cause them rather than by residential and small-commercial ratepayers; Commenter submits this comment 
in good faith and in furtherance of that public interest. Any terms relating to the adoption or use of the 
framework are matters Commenter is prepared to address in the appropriate forum and are not placed at 
issue in this comment. The technical invariants underlying this framework have been published openly 
and in full and made available for examination, comment, and challenge, consistent with the transparency 
these proceedings seek to advance. 
II. Documented Priority — Diagnostic-to-Action Correlation 
Commenter reported the specific conditions now before the Commission and the reliability community 
beginning January 17, 2026 — months before the corresponding federal actions. The following 
correlations are drawn from Commenter’s timestamped correspondence record (255 items, January 17 – 
June 29, 2026) and the public federal record. Each states the diagnostic, the date of first report, and the 
later federal action addressing the same condition. This documents temporal sequence, not causation. 
Diagnostic reported by Commenter 
First report 
Subsequent federal action 
Interval 
Sub-synchronous resonance / high-frequency “jitter” 
signature at data-center interconnects (4.83 MW 
harmonic) 
2026-01-17 
NERC Level 3 Alert (2026-05-04) — documents sub-
synchronous oscillation traced to computational 
facilities in ERCOT and Dominion territories 
107 d 
Inverter-Based Resource telemetry / high-speed (PMU) 
measurement gap for large loads 
2026-01-17 
NERC Essential Actions (2026-05-04) — high-speed 
monitoring at data centers made an Essential Action 
107 d 
Ghost Load — unmetered / behind-the-meter large-load 
extraction and metering bypass 
2026-01-17 
FERC show-cause (2026-06-18) — requires reform of 
co-location and BTM arrangements and accounting 
transparency 
152 d 
Cost-shifting of stranded / infrastructure costs onto 
residential ratepayers 
2026-01-17 
FERC (2026-06-18) — preliminarily finds large-load 
tariffs unjust and unreasonable; requires cost-recovery 
agreements 
152 d 
Sudden multi-hundred-MW computational load drop as 
grid-stability / blackout risk 
2026-01-20 
NERC Level 3 Alert (2026-05-04) — issued in 
response to repeated 1,000+ MW load-drop events 
104 d 
PJM / Northern Virginia (Dominion) corridor as point 
of first failure 
2026-01-20 
DOE 202(c) emergency orders (2026-06-30); NERC 
traces oscillation to Dominion territory (2026-05-04) 
104 d+ 
Need for a diagnostic / certification standard and 
generation-adequacy accounting for large loads 
2026-01-17 
FERC resource-adequacy reports (2026-06-18); NERC 
“Computational Load Entity” registration category 
(2026-05-04) 
107 d 
Priority intervals range 104–152 days. Commenter respectfully submits that this documented priority is relevant to 
the Commission’s evaluation of the methodologies proposed in response to its orders.

III. Technical Considerations for the Commission 
Commenter respectfully submits the following considerations, each corresponding to matters already 
before the Commission and the reliability community. 
A. From periodic review to continuous validation. 
Existing reliability mechanisms are predominantly retrospective — periodic reviews, seasonal 
assessments, and post-event investigations operating on intervals measured in weeks or quarters. Large 
computational loads change operating state on sub-second and sub-cycle timescales, driven by automated 
orchestration rather than human dispatch. A settlement or reliability process operating on a sixty-to-
ninety-day study cadence cannot resolve behavior that changes faster than the measurement interval; the 
relevant dynamics have completed before the review begins. Consideration should be given to whether 
reliability verification should migrate from periodic review toward continuously operating validation 
performed at the interconnection boundary, sampled at sub-cycle resolution, such that latent instability is 
detected in the telemetry stream before it propagates onto the Bulk Power System. 
B. Observability beyond consumption — high-resolution boundary telemetry. 
Conventional load modeling characterizes demand through aggregate consumption measured over 
settlement intervals. This resolution is insufficient to observe the behavior that large computational loads 
impose at the point of common coupling: sub-synchronous oscillation, harmonic injection, phase-
coherence deviation relative to the nodal reference frame, and rapid load-migration events. These 
signatures fall below the temporal resolution of SCADA-class metering and are only resolvable through 
Phasor Measurement Unit (PMU) and Dynamic Fault Recorder (DFR) instrumentation sampling at sub-
cycle rates. The NERC Essential Actions’ requirement for high-speed monitoring and dynamic fault 
recording at computational facilities identifies the same instrumentation gap. Consideration should be 
given to whether interconnection standards should require boundary telemetry of sufficient resolution to 
resolve the physical signature of computational load — the measurable fingerprint of its behavior at the 
grid interface — rather than telemetry describing only integrated historical consumption. 
C. Embedded reliability boundaries and machine-speed corrective action. 
Where load behavior changes at sub-cycle speed, corrective mechanisms that depend on human-initiated 
dispatch cannot respond within the window in which an oscillation or load-rejection event propagates. 
Consideration should be given to whether interconnection requirements should include automated, 
boundary-resident reliability functions — continuous invariant monitoring against defined tolerances, 
automated throttling or curtailment triggered when a measured quantity exceeds its bound, and machine-
speed corrective action — such that a deviation is arrested at the interconnection point before it becomes 
a system-wide event. This is a shift from retrospective compliance verification to continuous, automated 
enforcement of reliability bounds at the boundary. 
D. Execution sovereignty and software common-mode risk. 
Reliability planning has historically treated geographic and electrical diversity as sufficient protection 
against correlated failure. Large computational loads defeat that assumption at the software layer: 
thousands of megawatts distributed across multiple balancing authorities may nevertheless share common 
orchestration environments, model-serving stacks, or scheduling dependencies, such that a single 
upstream software event can command a synchronized change in demand across geographically dispersed 
facilities simultaneously. This is a common-mode failure mode invisible to redundancy analysis that

considers only physical topology. The documented record of sudden, synchronized reductions of large 
computational load — including the repeated 1,000-plus-MW load-drop events underlying the NERC 
Level 3 Alert — is consistent with a shared-dependency mechanism rather than independent local causes. 
Consideration should be given to whether interconnection study must characterize software and 
orchestration dependency — execution sovereignty — as a distinct common-mode risk alongside physical 
redundancy, and to whether the concentration of load behind shared control planes should be a studied 
parameter of interconnection. 
E. Dependency mapping and hidden burden. 
Traditional planning methodologies quantify capacity; future methodologies may need to characterize 
dependency — between facilities, orchestration systems, cloud environments, regions, and execution 
pathways — including which facilities share identical software assumptions, which execution pathways 
are coupled, and what percentage of computational activity could migrate simultaneously. Commenter’s 
Ghost Load™ methodology addresses burdens of this kind that remain hidden from present reliability and 
cost-allocation frameworks. The underlying question is significant whether or not this terminology is 
adopted: whether operators can reliably govern, and correctly allocate the cost of, dependencies that 
cannot yet be observed. 
F. Cost-causation. 
The reliability burdens created by large computational loads — including the burdens arising from the 
observability, validation, and dependency considerations above — should be assigned to the loads that 
cause them, on a cost-causation basis, rather than socialized onto residential and small-commercial 
ratepayers. The present record — including PJM capacity auction results, Independent Market Monitor 
findings, the NERC Level 3 Alert, and documented instances of cost-shift in transmission planning — 
supports the need for cost-allocation methodologies that preserve reliability invariants at the point of 
interconnection while assigning burden on a cost-causation basis. 
G. Curtailment priority and observability of federally-accelerated computational load. 
In its Section 202(c) emergency orders during the June–July 2026 heat events (e.g., DOE Order No. 202-
26-32, June 30, 2026, and its extensions through July 7, 2026), the Department of Energy directed 
curtailment of data centers and large loads of 50 MW or greater onto backup generation within fifteen 
minutes of an emergency signal, while expressly exempting hospitals, 911 call centers, water treatment 
plants, air traffic control facilities, and defense facilities. Separately, the Department’s Genesis Mission 
(Executive Order, November 24, 2025) is federally accelerating large computational load and expressly 
targets grid interconnection, planning, and operations, with a stated national-security dimension. 
Consideration should be given to how curtailment priority is to be ordered where federally-directed or 
security-associated computational load may qualify for exemptions available to critical facilities — such 
that it could sit ahead of, rather than behind, ordinary large loads in a curtailment sequence — and to how 
the Commission’s cost-causation, transparency, and observability objectives are to be preserved with 
respect to computational load whose operating characteristics may be shielded from the observability 
these proceedings seek to establish. Commenter raises these as questions the record should resolve and 
takes no position on any federal program as such. 
H. Unpriced resource consumption and diffuse externalized costs.

Large computational loads impose costs beyond the price of energy — costs that, because they are diffuse 
and fall across many parties, are difficult to trace and are frequently borne by the public rather than by the 
causing load. Consideration should be given to whether the Commission’s cost-allocation methodologies 
account for, and assign on a cost-causation basis: (i) water consumed for compute cooling, including non-
recycled evaporative consumption and its effect on water availability for other public needs, including 
firefighting; (ii) fire and air-quality risk associated with backup-generator operation during curtailment 
events, including operation beyond ordinary emission limits authorized under emergency orders; (iii) 
accelerated wear on transmission, distribution, and related public infrastructure from large-load stress and 
rapid load-switching; (iv) casualty and reliability risk as reflected in rising insurance costs and coverage 
disputes affecting facilities and surrounding communities; and (v) the administrative and adjudicative 
burden — borne by public agencies, ratepayer advocates, and the courts — of resolving the proliferating 
disputes among utilities, States, insurers, and customers over who bears these costs. As independent 
analysis has observed, the confidential nature of many utility–large-load arrangements makes the public’s 
share of these costs difficult to measure; that opacity itself supports the need for the transparency and 
cost-causation discipline these proceedings contemplate. 
I. Transparency and the orientation of the just and reasonable standard. 
The Commission’s authority under the Federal Power Act was established to protect consumers and to 
ensure that rates are just and reasonable, and cost causation is the foundation of that standard. That 
standard requires that any determination rest on a rational connection between the facts found and the 
conclusion reached, on the evidence before the Commission. A structural difficulty is presented where the 
same federal authority both directs the Commission to ensure transparency, cost-causation discipline, and 
ratepayer protection with respect to large computational loads, and, through separate mechanisms — 
contractual confidentiality, emergency-order expedience, and national-security classification associated 
with federally-directed computational load — withholds from the Commission the information those 
objectives require. Consideration should therefore be given to a threshold question: whether the 
Commission can discharge its statutory duty, owed to the public the Act was enacted to protect, with 
respect to costs, risks, and load behavior it is structurally prevented from observing. Where the public 
bears the costs, absorbs the reliability risk, and finances the infrastructure of large computational load, yet 
is denied — through confidentiality, emergency expedience, or classification — the transparency 
necessary to verify how those costs are caused and allocated, consideration should be given to whether 
the arrangement serves the public the Commission was created to protect, or inverts that relationship, 
such that the public is made to serve the institutions rather than the institutions serving the public. 
Commenter raises this as a consideration for the Commission’s weighing, and asserts no legal conclusion.

[No extractable text on this page.]

IV. Toward a Resolution — Verification at the Point of Interconnection 
The considerations above describe a common difficulty: the Commission is directed to ensure just and 
reasonable rates, reliability, and cost-causation for large computational loads whose behavior and costs it 
may be structurally prevented from observing. Commenter respectfully submits that a resolution exists 
that does not require the Commission, the Regional Transmission Organizations, or the operators to 
obtain access to the proprietary or protected interior of any facility. Consideration should be given to 
whether a standard of the following character — verification performed at the electrical boundary, the 
point of common coupling, rather than within the facility — would resolve the difficulties identified 
above. 
Such a standard would operate as a sequence of verification gates applied at interconnection, each 
addressing a specific consideration raised in this comment: 
Gate I — Boundary diagnostic (timing and stability). 
The first gate establishes the trustworthiness of the telemetry itself and the timing stability of the load at 
the point of common coupling. Using PMU and Dynamic Fault Recorder data sampled at sub-cycle 
resolution, it measures phase coherence — the rotational alignment of the load’s injected and drawn 
power relative to the nodal reference frame — together with sub-synchronous oscillation and harmonic 
signature. Two invariants govern this stage: a bounded phase-jitter tolerance, measured against a 
reference clock, which sets the maximum permissible timing deviation of the interconnection; and a 
bounded controller-response interval, which sets the maximum permissible latency of the load’s control 
response. Together these determine whether the telemetry stream is coherent enough to be relied upon and 
whether the load’s timing behavior is within stable bounds. This gate supplies the high-resolution 
observability the reliability record identifies as absent (Considerations A and B), measured entirely at the 
boundary and without access to internal facility operations. 
Gate II — Boundary dependency map. 
The second gate characterizes the load’s coupling and dependency profile as observed at the boundary: its 
electrical relationship to proximate generation and to other loads, the load-migration fraction — the 
proportion of demand capable of shifting simultaneously — and the degree to which the facility’s 
behavior is correlated with shared control planes rather than independent local dispatch. This maps the co-
dependency of electrically proximate load and generation that the reliability record now requires be 
studied together, and supplies the data needed to assess the software common-mode exposure described in 
Considerations D and E — derived from observed boundary behavior rather than from disclosure of 
internal architecture. 
Gate III — Boundary settlement (throughput reconciliation and cost-causation). 
The third gate reconciles declared load against delivered load and provides the measurable basis for cost-
causation. Two invariants govern this stage: a throughput differential — the measured difference between 
the load authorized or represented and the load physically drawn and delivered at the boundary — and an 
operational bound, the tolerance within which that differential is treated as accounted-for. A differential 
exceeding the bound identifies unaccounted or displaced throughput: service received but not reflected in 
the declared or metered figure. This is the physical basis of the cost-shift and unmetered-extraction 
concerns before the Commission (Considerations F and H). Because the differential and its bound are

computed from boundary telemetry that the operator already records at settlement points, cost-causation is 
established from the measured record rather than from confidential internal accounts. 
Gate IV — Boundary certification (integrated operational alignment). 
The fourth gate is not a single measurement but a certification of the integrated result of Gates I through 
III. It certifies, against the full set of boundary invariants, whether the load can ramp, curtail, and operate 
flexibly while remaining within the accepted stability envelope — that is, whether its timing (Gate I), its 
dependency profile (Gate II), and its throughput reconciliation (Gate III) together place it inside the 
bounds required for reliable interconnection. Certification attaches to the measured, integrated boundary 
state of the load, not to representations about internal intent, and is continuous rather than one-time: a 
load that drifts outside the envelope fails certification and is subject to the automated corrective action 
described in Consideration C. This gate provides the operational-alignment and flexible-service 
determination the Commission’s reform categories contemplate, expressed as a verifiable state rather than 
a representation. 
The resolution of the transparency difficulty. 
Because each gate operates at the electrical boundary rather than within the facility, this class of standard 
resolves the difficulty described in Consideration I without disturbing any legitimate interest in 
confidentiality, emergency response, or classification. The interior of the facility — its proprietary 
methods and any classified activity — remains undisclosed. What becomes observable and verifiable is 
only the load’s behavior at the point where it touches the public grid: its timing, its dependency, its cost-
causation, and its stability. The Commission thereby obtains the evidentiary basis its just and reasonable 
standard requires, and the operators obtain observability at the speed of the load, while the facility’s 
internal integrity is preserved. Consideration should be given to whether a boundary-verification standard 
of this character should be incorporated into the interconnection and cost-allocation methodologies these 
proceedings will produce. Commenter has developed, and holds prior-art and pending intellectual-
property rights in, a specific and complete embodiment of a standard of this character, as set forth in the 
Reservation of Rights below, and is prepared to describe that embodiment in full in response to the 
Commission’s inquiries. 
V. Relationship to the Current Record and Pending Deadlines 
The considerations above correspond to matters already before the Commission and the reliability 
community. The following summarizes the recent federal record and the two imminent deadlines that will 
develop it, illustrating that the reliability conditions at issue are active and that the record is presently in 
motion: 
Date 
Matter 
2026-05-04 
NERC Level 3 Essential Actions Alert — modeling, instrumentation, and high-speed monitoring of computational loads; 
sub-synchronous oscillation traced to Dominion (Northern Virginia) territory. 
2026-06-18 
FERC Section 206 show-cause orders to all six RTO/ISOs; large-load tariffs preliminarily found unjust and unreasonable; 
five reform categories including cost-shifting and transparency. 
2026-06-30 
DOE Section 202(c) emergency orders (Order No. 202-26-32) directing PJM to curtail large loads onto backup generation; 
Northern Virginia corridor flagged amid record demand. 
2026-08-03 
NERC Level 3 Alert response deadline — registered entities must report status on the seven Essential Actions (33 
questions). 
2026-08-17 
FERC show-cause / tariff-reform deadline (60 days from June 18) — each RTO/ISO must show cause or file reforms on 
large-load integration.

Commenter notes that the substance of these considerations — including intent-aware observability, 
continuous validation, software common-mode exposure, and dependency-based cost allocation — was 
authored and publicly timestamped prior to the actions now under consideration, and respectfully submits 
that this priority is relevant to the Commission’s evaluation of methodologies proposed in response to its 
orders. 
VI. Reservation of Rights 
Commenter has authored and publicly timestamped a proprietary diagnostic, observability, and stability 
framework and holds pending United States Patent and Trademark Office applications relating thereto 
(USPTO Serial Nos. 99598875, 99600821, 99613073, 99717240, 99729215, and 99745529). To the 
extent any methodology adopted, referenced, or incorporated in these proceedings — or in any Regional 
Transmission Organization, Independent System Operator, or Transmission Owner filing responsive to 
the Commission’s orders — incorporates, derives from, or substantially approximates that framework or 
its associated marks, Commenter reserves all rights, including the right to seek appropriate relief in the 
proper forum. This comment is submitted to place the Commission and the participants in these dockets 
on notice of Commenter’s claimed proprietary interest and reservation of rights, and not to assert any 
claim for adjudication within these proceedings. 
VII. Conclusion 
For the foregoing reasons, Commenter respectfully requests that the Commission consider these technical 
considerations in the above-captioned proceedings and take note of the reservation of rights stated herein. 
Commenter offers these considerations in good faith and in the shared interest of a reliable Bulk Power 
System and of just and reasonable rates for the public the Commission was established to protect. 
Respectfully submitted, 
_______________________________ 
Lisa Melton aka L.M. Marlowe 
lm.marlowe@pm.me 
Intervenor / Commenter, pro se 
Related federal references: GAO No. COMP-26-002174; DOE Administrative Claim No. AR 2026-001.