The 8760 Withdrawal Study: How Hourly Grid Headroom Modeling and SAM-Based BESS Sizing Unlock Large Load Interconnection in SPP’s HILLGA and Every Other ISO

By Keentel Engineering — Power System Studies | EMT & Dynamic Modeling | Interconnection Support

Every large load project — every AI data center, crypto campus, electrolyzer, and industrial expansion — eventually collides with the same question, and it is not “will the utility approve us?” It is sharper than that:

How much power can this site actually withdraw from the grid hour by hour, across a real year of system conditions?

That number is not the load’s nameplate. It is not the substation’s transformer rating. It is not a constant. It moves with seasonal facility ratings, system load, generation dispatch, and outages — and at most points of interconnection in America today, it is smaller than developers hope. The gap between what a site wants to draw and what the grid can deliver in its worst hours is precisely why behind-the-meter batteries and on-site generation have become the defining architecture of the AI-era data center.

But you cannot size a battery against a number you do not know. And you should not file an interconnection application around one either.

This article explains the 8760 withdrawal study an hour-by-hour model of grid headroom at your POI — and how Keentel Engineering pairs it with NREL’s System Advisor Model (SAM) to size battery energy storage with national-laboratory-grade rigor. We then walk through how this analysis plugs directly into SPP’s new HILL/HILLGA framework and the interconnection processes of every other major ISO. A detailed FAQ follows.

Part 1: What an 8760 Withdrawal Study Actually Is

“8760” is the number of hours in a year — and an 8760 withdrawal study produces one number for each of them: the maximum MW the site can withdraw from the transmission system in that hour without causing a reliability violation.

Getting there takes three layers of engineering:

Layer 1 — The Shadow Load Study (SLS)

Before anything else, we replicate the analyses the RTO and transmission owner will eventually perform on your application, using the same class of models and methodology they use: AC power flow with thermal and voltage screening at the POI, the applicable NERC TPL-001-aligned contingency set, short circuit review, and system-strength screening (in SPP, consistent with the SCRCCT methodology and its SCR/WSCR/CSCR ≥ 6.0 and CCT ≥ 0.15 s thresholds). The output is foresight: the expected study outcome, the binding constraints, the likely network-upgrade exposure, and the maximum supportable withdrawal by season — known before deposits are paid and the study clock starts.This initial assessment draws on comprehensive power system studies to identify thermal, voltage, stability, and system-strength constraints at the POI.

Layer 2 — The hourly mapping

Grid capability is not one number per season; it is a distribution. We build a representative-state matrix — season × system load level × outage condition × dispatch pattern — solve the withdrawal limit for each state, then map every one of the 8,760 hours of the year to its state using multi-year historical system load, seasonal ratings calendars, outage seasonality, and weather. The result is an hourly withdrawal-capacity time series, with every hour traceable to its governing constraint.

Layer 3 — Deficit analysis

Overlay the facility’s hourly demand profile, and the deficits fall out: which hours the grid comes up short, how deep, how long, how often, and how they cluster. The duration curve and the deficit statistics are the raw material of every design decision that follows — most importantly, the battery.

Part 2: BESS Sizing in SAM Why We Use a National Laboratory’s Tool

Once the deficit hours are known, the battery question sounds simple: big enough to cover them. In practice, naive sizing fails in the details — clustered back-to-back deficit events that a battery cannot recharge between, round-trip losses, degradation over project life, temperature effects, and the subtle trap that recharging the battery is itself grid withdrawal that must fit inside the same hourly headroom.

That is why Keentel performs BESS sizing in NREL’s System Advisor Model (SAM) — the U.S. Department of Energy’s free, open-source techno-economic modeling platform, developed at the National Renewable Energy Laboratory with Sandia, used by more than 130,000 users in 190+ countries, launched somewhere in the world every two minutes, and relied on commercially by utilities and developers including AEP, Southern Company, and EPRI. NREL describes SAM’s battery model as the only publicly available tool of its kind accounting for voltage characteristics and both calendar and cycle degradation — with automated dispatch strategies for behind-the-meter and front-of-meter operation.

Our workflow connects transmission engineering to storage design in one auditable chain:

1. PSS®E produces the constraint — the 8760 withdrawal-capacity series from Layers 1–2, built on the RTO’s own models and contingency definitions.
2. SAM consumes it — configured as a behind-the-meter battery system: the data center’s hourly demand as the load profile, the on-site PV array modeled from site weather data, the lithium-ion battery with degradation and temperature models enabled, and the withdrawal series applied as the hourly grid import limit. SAM then dispatches the battery through all 8,760 hours: discharging when demand exceeds grid headroom plus PV, recharging from PV surplus and from the grid only within remaining headroom.
3. PySAM automates the search — scripted parametric sweeps across battery power and energy candidates, harvesting unmet-load hours, cycling, and end-of-life capacity, to identify the minimum-size frontier that achieves zero unmet load or the client’s chosen reliability target. Weather-year ensembles support probabilistic (P50/P90-style) sizing where the client wants reliability stated statistically.
4. Economics ride along for free — with the applicable utility rate structure loaded (from the OpenEI Utility Rate Database or the client’s contract terms), the same simulation quantifies demand-charge and energy-cost impacts of the dispatch, turning a sizing study into an operating pro forma input.

The deliverable is not “we recommend 150 MW / 600 MWh.” It is: here is the hourly headroom at your POI with constraint-level traceability; here is the deficit structure; here is the sizing frontier from a national-lab-validated simulation; here is the recommended size with margin, its sensitivity to weather years and load phasing, and its operating economics. That is a document a lender’s independent engineer can audit — and that an RTO study engineer will recognize as built on their own methodology.

Part 3: How This Unlocks SPP’s HILL and HILLGA Processes

SPP’s High Impact Large Load framework — the HILL Delivery Point Study (HDPS) for loads and the HILLGA generator assessment (Attachment BB) for their dedicated generation — is the newest, fastest large-load pathway in any U.S. RTO. It is also a framework in which the withdrawal study is not merely useful but load-bearing:

It sets the application’s requested MW intelligently. The HDPS clock is 90 days — but only after a complete, validated submittal, and the results are only as good as the MW you request. Oversubscribe a weak POI and the study returns upgrade costs and timelines that kill the pro forma; undersubscribe and you strand capacity. The SLS tells you the answer before SPP does.

It de-risks SCRCCT screening. A 500 MW power-electronic load at a rural POI faces real probability of failing SPP’s system-strength screen, which triggers supplemental EMT studies outside the 90-day window. The withdrawal study’s SCR/WSCR/CSCR and CCT screening quantifies that risk site by site — informing both siting decisions and whether to invest in a submission-ready PSCAD model (which SPP itself recommends providing at application).

It feeds the HILLGA sizing formula and the LLRIS envelope. Under Attachment BB, supporting generation may be sized up to the HILL’s requested MW times the higher of 110% plus the seasonal Planning Reserve Margin or 125% — and once operating, the generator’s injection is tied to the load under Load Limited Resource Interconnection Service. Both the sizing decision and the LLRIS operating envelope are functions of the load’s forecast and the grid’s hourly capability. The 8760 study is, in effect, the quantitative backbone of the HILL ten-year forecast, the generator sizing case, and the BESS that firms the difference.

It converts curtailment-style service into designable numbers. SPP’s conditional pathways trade speed for potential curtailment during system stress. Whether that trade is acceptable is not a philosophy question — it is a deficit-hours question. An hourly headroom model prices the curtailment exposure directly, and sizes the storage that neutralizes it.



For multi-site portfolios, it is the ranking engine. Developers evaluating several candidate sites — increasingly the norm for powered-land strategies — can compare hourly headroom, deficit depth and duration, upgrade exposure, system strength, and required BESS size on one normalized matrix, and put capital where the electrons are.

Part 4: The Same Analysis Travels to Every Other ISO

The tariff names change; the physics does not. Every organized market now confronts the same collision between hyperscale load growth and finite transmission headroom, and each is building processes where hourly withdrawal modeling pays:

ERCOT

The original large flexible load market. ERCOT’s interconnection framework for large loads — with its emphasis on load flexibility, curtailability, and (for co-located configurations) netting arrangements — is fundamentally an hourly-headroom conversation. An 8760 model quantifies exactly how much firm service a site can count on, what a curtailable arrangement is actually worth, and what storage converts flexible access into firm compute capacity. Notably, SPP expressly drew on ERCOT’s large electronic load ride-through philosophy in writing its HILL requirements — experience in one footprint transfers to the other, and Keentel’s large-load practice spans both, from 50 MW to 1 GW engagements.

CAISO and the California utilities

Load interconnects through the distribution/transmission planning processes of the IOUs (Rule 21 territory and wholesale distribution access), where hosting capacity, deliverability, and grid-code compliance dominate. The same headroom-and-storage logic applies at the POI, and California’s rate structures make the SAM economic layer especially valuable — demand charges and TOU spreads materially change optimal BESS dispatch.

PJM and MISO

Both are actively reforming large-load interconnection amid unprecedented data center queues. Load additions flow through transmission-owner load interconnection procedures and RTO planning review; co-located load-at-generation configurations have drawn intense FERC attention in PJM. In either footprint, arriving with an independent withdrawal analysis — built on the RTO’s planning models and TPL-aligned contingencies — collapses study surprise and positions the BESS as a designed mitigation rather than an afterthought.

NYISO and ISO-NE

Smaller footprints, tighter systems, and aggressive state clean-energy overlays make headroom scarcer and the storage case stronger; the methodology is identical.

Across all of them, the through-line is this: the RTO’s study tells you what the grid will give you; the withdrawal study tells you first — and the SAM layer turns the shortfall into a designed, priced, financeable storage solution. Keentel performs this stack with the same disciplines in every market: the RTO’s own models where accessible, TPL-001-aligned contingency frameworks, system-strength screening, and NREL-toolchain storage design, all under the responsible charge of a licensed Professional Engineer.

Part 5: What a Keentel 8760 Withdrawal & BESS Study Includes

• Shadow Load Study report — binding constraints, indicative upgrade exposure, seasonal maximum-withdrawal table, short circuit review, and system-strength screening with supplemental-study risk rating.
• Methodology alignment memo — a mapping of every analysis to the corresponding RTO/TO study element, so results are directly comparable to the official study when it arrives.
• 8760 withdrawal-capacity time series (Excel/CSV) with hour-level traceability to governing constraints, plus duration curves and deficit statistics.
• SAM-based BESS sizing — the feasible-size frontier, recommended power/energy with engineering margin, weather-year and load-phasing sensitivities, degradation-aware end-of-life performance, and dispatch economics under the applicable rate structure.
• Portfolio comparison matrix (multi-site programs) — normalized site ranking for capital allocation.
• Application-ready outputs — the forecast, single-line, and load modeling data (CMLD/PERC1) that flow directly into the interconnection submittal, with PSCAD EMT models available where system-strength screening warrants.
  

Work With Keentel Engineering

Keentel Engineering provides the complete withdrawal-study stack — Shadow Load Studies, 8760 withdrawal-capacity modeling, SAM-based BESS sizing, and the application, dynamic modeling (CMLD/PERC1), and PSCAD EMT deliverables that follow — for large loads across SPP, ERCOT, CAISO, PJM, and every North American market.

Evaluating sites? Preparing an application? Staring at a study report you didn’t expect? Contact us at contact@keentelengineering.com or 813-389-7871, or visit keentelengineering.com to schedule a technical consultation.

Our Offices: - Head Office – Tampa, FL: 400 N Ashley Dr, STE #2600, Tampa, FL 33602 · (813) 389-7871 - Austin, TX: 5900 Balcones Drive, STE 100, Austin, TX 78731 · (512) 591-0752 - Sacramento, CA: 1401 21st St, Ste R, Sacramento, CA 95811 · (916) 913-4524 - Baltimore, MD: 306 W Redwood St, STE 200, Baltimore, MD 21201 · (410) 225-2181

This article reflects SPP’s HILL/HILLGA framework (RR696/RR720/RR724, Attachment BB, BP 7250/7850)

Frequently Asked Questions