GridArb is a fast first check, not a bank-ready study. It tells you which sites are worth a $75,000 study deposit and a consultant, and which to drop today.
PSEG zone, PJM, day-ahead hourly prices, 2025-01-01 to 2025-12-31. Perfect-foresight energy arbitrage for a 1 MW / 2 MWh battery.
Data as of 2026-10-12 (latest day-ahead price in the database). Every figure carries a numbered source; see Sources at the end.
Annualised. Energy arbitrage comes from the backtest above (day-ahead hourly prices). Capacity uses the PJM Base Residual Auction clearing price of $325.00/MW-day[5,7], ELCC 0.59 for a 4-hour class on 0.5 MW effective[21]. Rows marked included make up the base total.
| Revenue line | $/year | $/kW-year | Note |
|---|---|---|---|
| Energy arbitrage, perfect foresight (ceiling)[1,23] | $18,508 | 18.51 | Optimizer sees each day's prices in advance; day-ahead hourly prices. |
| Energy arbitrage, achievable (x0.80)[12] | $14,806 ($12,956 to $16,657) | 14.81 (12.96 to 16.66) | Perfect foresight x 0.80; range shown for x0.70 to x0.90. |
| Capacity revenue, base ELCC 0.59 (2028/29)[5,7,21] | $34,994 | 34.99 | 0.5 MW effective x ELCC 0.59 = 0.295 MW UCAP x $325.00/MW-day x 365. |
| Capacity revenue, low-ELCC scenario (0.40)[5,8] | $23,725 | 23.73 | Sensitivity only: PJM long-range ratings for 4 h storage trend toward 0.38-0.42. |
| Regulation (PJM frequency regulation), forward base case from COD 2029 | $45,990 ($3,139 to $213,352) | 45.99 (3.14 to 213.35) | Forward path from the first operating year 2029 (COD): base $35.00/MW-h ($150 in 2026 falling to $35 by 2029, then flat), low $4.30 (pre-redesign average to the $4.30 cost floor by 2028), high $81.18 (last-12-month average falling to $35 by 2033, never below base; Modo Energy, PJM regulation market outlook (2026)). Cleared-hours share in 2029: 10% / 15% / 30%, from PJM's 750 MW requirement over an expected battery fleet of 5,000 MW (low 1.5x, high 0.5x that fleet; capped at 80%), falling as the fleet grows; in the idle 20 hours a day and, where it beats arbitrage, the cleared share of the active hours (that share of the arbitrage is forgone). RegUp / RegDn split 50/50: a placeholder until PJM publishes the two prices. Context only: what the last 12 months paid if you had cleared every hour: $1,014,601 a year at $115.82/MW-h (data through 2026-10-13). Regulation prices are expected to fall as more batteries join PJM: the market is a few hundred MW, so new storage moves it quickly. Treat this line as upside, never as the base for a loan. |
| NJ GSESP incentive (Phase 1, 15 years)[9,10] | see note | No GSESP route today: program requires >= 5 MW. | |
| TOTAL base case, expected from COD 2029 (achievable arbitrage + capacity + regulation, base ELCC)[12,7] | $93,570 ($51,089 to $260,006) | 93.57 (51.09 to 260.01) | First operating year 2029. Excludes GSESP; regulation wins 15% of the active hours (the cleared share), so that part of the arbitrage line is left out of this total. |
Regulation. Forward path from the first operating year 2029 (COD): base $35.00/MW-h ($150 in 2026 falling to $35 by 2029, then flat), low $4.30 (pre-redesign average to the $4.30 cost floor by 2028), high $81.18 (last-12-month average falling to $35 by 2033, never below base; Modo Energy, PJM regulation market outlook (2026)). Cleared-hours share in 2029: 10% / 15% / 30%, from PJM's 750 MW requirement over an expected battery fleet of 5,000 MW (low 1.5x, high 0.5x that fleet; capped at 80%), falling as the fleet grows; in the idle 20 hours a day and, where it beats arbitrage, the cleared share of the active hours (that share of the arbitrage is forgone). RegUp / RegDn split 50/50: a placeholder until PJM publishes the two prices. Context only: what the last 12 months paid if you had cleared every hour: $1,014,601 a year at $115.82/MW-h (data through 2026-10-13). Regulation prices are expected to fall as more batteries join PJM: the market is a few hundred MW, so new storage moves it quickly. Treat this line as upside, never as the base for a loan.[13,14]
Benchmark check: First operating year 2029: GridArb base case $93,570 per MW-year against Modo Energy's post-redesign PJM fleet average of $744,000 per MW-year (0.13x). 15-year average: $67,108 per MW-year against Modo's central case of $84,000 per MW-year for a 4-hour PJM battery, 2027-2049 (0.80x; the warning fires above 1.5x).[15]
Federal ITC: 30% assumed (Section 48E), FEOC content rules apply. Cost side only; not modelled in detail.[11]
| Delivery year | RTO clearing price $/MW-day | ELCC 4 h | 6 h | 8 h | 10 h | Status |
|---|---|---|---|---|---|---|
| 2026/27 | 329.17 | 0.50 | 0.58 | 0.62 | 0.72 | cleared at the cap |
| 2027/28 | 333.44 | 0.58 | 0.67 | 0.70 | 0.78 | cleared at the cap |
| 2028/29 | 325.00 | 0.59 | 0.68 | 0.71 | 0.78 | cleared at the cap (used in this report) |
| 2029/30 | collar-bounded, auction December 2026 | last year under the price collar |
This battery is 2 hours, under the 4-hour class: it is counted as 0.5 MW effective (energy / 4) at the 4-hour ELCC. The owner will confirm this rating in PJM's ELCC calculator before it is used in a report.[7,19]
The FERC-approved price collar ends after 2029/30; later prices are unknown (PJM's uncapped simulations ran $530 to $555/MW-day) and the fifteen-year view labels them as scenarios.[6]
Transmission-connected (front-of-meter). Capacity LDA: PSEG (PS-North), inside EMAAC. Transmission-connected: eligible for capacity revenue once the PJM interconnection process grants Capacity Interconnection Rights (CIRs). PJM queue timing and network-upgrade costs are not estimated here and can kill a project.[22]
| Month | Days | Optimizer | Naive | Opt / naive | Cycles | Best day | Best day $ |
|---|---|---|---|---|---|---|---|
| 2025-01 | 31 | $1,910 | $775 | 2.47 | 29.3 | 2025-01-20 | $210 |
| 2025-02 | 28 | $1,067 | $337 | 3.16 | 25.7 | 2025-02-21 | $130 |
| 2025-03 | 31 | $1,228 | -$140 | n/a | 29.8 | 2025-03-03 | $97 |
| 2025-04 | 30 | $1,030 | -$156 | n/a | 28.6 | 2025-04-09 | $106 |
| 2025-05 | 31 | $619 | $221 | 2.79 | 26.3 | 2025-05-16 | $77 |
| 2025-06 | 30 | $2,191 | $2,090 | 1.05 | 26.5 | 2025-06-24 | $591 |
| 2025-07 | 31 | $4,110 | $3,961 | 1.04 | 29.4 | 2025-07-29 | $803 |
| 2025-08 | 31 | $1,215 | $1,154 | 1.05 | 27.1 | 2025-08-12 | $143 |
| 2025-09 | 30 | $938 | $807 | 1.16 | 27.2 | 2025-09-29 | $66 |
| 2025-10 | 31 | $1,593 | $952 | 1.67 | 30.2 | 2025-10-06 | $94 |
| 2025-11 | 30 | $1,152 | $882 | 1.31 | 27.6 | 2025-11-03 | $83 |
| 2025-12 | 31 | $1,441 | $939 | 1.53 | 27.9 | 2025-12-16 | $169 |
| Total | 365 | $18,495 | $11,823 | 1.56 | 335.6 | 2025-07-29 | $803 |
Net of degradation. Blue bars: optimizer. Orange line: naive schedule (charge 01-05, discharge 17-21 local).
Optimizer $803 vs naive $822. Hours are America/New_York.
Default 1 MW / 2 MWh battery, cached results. * marks the best $/kW-year. Shaded row is this report's run when it matches.
| Zone | Market | Year | Total | $/kW-yr | Opt / naive | Spread $/MWh | Cycles |
|---|---|---|---|---|---|---|---|
| AECO | DA | 2024 | $11,443 | 11.42 | 1.44 | 23.2 | 314 |
| AECO | DA | 2025 | $18,747 | 18.76 | 1.58 | 33.0 | 334 |
| JCPL | DA | 2024 | $10,850 | 10.83 | 1.45 | 22.2 | 314 |
| JCPL | DA | 2025 | $18,512 | 18.52 | 1.60 | 32.7 | 334 |
| PSEG | DA | 2024 | $10,842 | 10.82 | 1.44 | 22.3 | 314 |
| PSEG | DA | 2025 | $18,495 | 18.51 | 1.56 | 32.6 | 336 |
| RECO | DA | 2024 | $11,566 | 11.54 | 1.39 | 23.2 | 318 |
| RECO | DA | 2025 | $19,798 | 19.81 | 1.56 | 34.1 | 340 |
Base case 49.80 $/kW-year (achievable arbitrage + capacity, base ELCC). Rows are ordered by size of impact.[38]
| If this moves | Base | Alternative | Total base case $/kW-year | Change | Note |
|---|---|---|---|---|---|
| ELCC 0.40 (low scenario) | 0.59 | 0.40 | 38.54 | -11.26 | |
| Capacity price -30% | $325.00/MW-day | $227.50/MW-day | 39.30 | -10.50 | |
| Degradation $10/MWh | $5/MWh | $10/MWh | 47.32 | -2.48 | |
| Cycles per day 0.8 | 1 | 0.8 | 47.46 | -2.34 | approximation from cached daily results (conservative); not re-optimised |
| Achievable factor 0.70 | 0.80 | 0.70 | 47.95 | -1.85 | |
| Achievable factor 0.90 | 0.80 | 0.90 | 51.65 | +1.85 | |
| Round-trip efficiency 0.83 | 0.87 | 0.83 | 48.64 | -1.16 |
Connection score: Green: no screening flag.
GridArb screening thresholds (assumptions): feeder kW vs planned kW at 1.0 and 0.5; substation within 3 and 10 km; queue 1 or 3 at the same substation, 3 or 8 within 5 miles; upgrades $150k and $500k per MW.
What only a study can tell you
Local rules: Town not identified from the address; check the municipal zoning code and any moratorium before you spend. GridArb tracks NJ battery-storage moratoriums and bans by hand (1 active municipality, last checked 2026-10-27).[36] Tracker.
Site: no address given
Jurisdiction and process: PJM cycle route (69 kV and above) (from the connection type 'transmission' (no voltage given; add voltage_kv=...); the line is 69 kV). PJM Generation Interconnection, cycle process (Manual 14H): application window, Phase 1 to 3 system impact studies, Interconnection Service Agreement, then construction.[32,17]
Deposits and fees. PJM study deposit plus the readiness deposit per MW (Manual 14H; figures for 0-20 MW below).
Timeline. Typical timeline under the cycle process: about 2 to 3 years from the application window to an interconnection agreement, then construction. Small (under 20 MW) projects can use the expedited process when the system-impact screens pass. These are typical durations, not PJM commitments.
Hosting capacity. Enter a site address or lat/lon to read the utility map, or the circuit headroom by hand. Map: PSE&G.[28]
Nearest substations: Enter a site address or lat/lon to look up nearby substations. (add address=... or lat=...&lon=... to the report URL).[27]
PJM queue, active cycle projects with this zone's transmission owner: 4 active of 4 listed; list dated 2026-10-28. PJM's public cycle queue list has no coordinates; projects are matched to the OpenStreetMap substations around the site by name and placed at that substation's distance. Unmatched projects in the zone are counted, not placed.[29]
At the nearest substation: none matched.
Within 5 miles: none matched.
5 to 10 miles: none matched.
4 active project(s) in the zone could not be placed by name.
Network-upgrade cost per MW allocated to recent projects in this zone: $7,788 per MW ($38,403,000 over 4,931 MW; 2 of 4 projects carried an allocation; TC2 PH2). A warning signal, not your cost: PJM's studies allocate by impact.[30]
PJM Queue Scope. Queue Scope is PJM's free evaluator of a point of interconnection against the queue: open it (PJM account needed), pick the state and transmission owner, choose the substation or line you plan to connect to and the MW, run the evaluation, and export the result to Excel. Attach that export to your GridArb site report and the reviewer reads it with your numbers. Open Queue Scope.[31]
On a GridArb site report you can attach your Queue Scope export to the private report page.
Request a pre-application report PSE&G, default fee $300. NJ's PAVE pre-application process may not be live at every utility yet; where it is not, send the request to the utility's interconnection contact.
Connection type: transmission. Capacity LDA: PSEG (PS-North), inside EMAAC.
PJM study deposits. Study deposit $75,000 plus a readiness deposit of $4,000/MW ($4,000 for 1 MW): $79,000 to enter the queue. PJM Manual 14H, Generation and Transmission Interconnection Process.[17]
Your interconnection estimate. none entered (add interconnection_cost=... to the report URL); the cost and payback below assume $0 for upgrades.
What GridArb cannot know:
The utility's own interconnection study decides. Network-upgrade costs and PJM queue timing are not estimated here and can kill a project.
Tax credit (your choice): 30% ITC No ITC (fails the foreign-entity rule) Default 30%; a project owned by or materially helped by a prohibited foreign entity gets no credit.[37]
Benchmark costs, not quotes. Costs are the NREL ATB utility-scale benchmark plus a small-project premium unless you enter a quote (query parameters cost_power_kw and cost_energy_kwh; itc=false removes the credit; discount_rate sets the rate). Degradation and augmentation are a simple linear assumption[26].
Augmentation year 10. Total net cash over the horizon -$220,244.
Cost sensitivity, battery cost +/-20% (the interconnection estimate is held fixed).[20]
| Build cost | Capex after ITC | Simple payback | NPV |
|---|---|---|---|
| Cost -20% | $534,108 | 13.8 years | -$183,873 |
| Cost as shown | $667,635 | not within the horizon | -$392,465 |
| Cost +20% | $801,162 | not within the horizon | -$601,057 |
Import tariff scenario, not a forecast. Battery cost +0% / +25% / +80% (the interconnection estimate is held fixed); tariff rates on cells and packs have changed several times and may again.[35]
| Import tariff scenario | Capex after ITC | Simple payback | NPV |
|---|---|---|---|
| Battery cost +0% (no new tariff) | $667,635 | not within the horizon | -$392,465 |
| Battery cost +25% | $834,544 | not within the horizon | -$653,205 |
| Battery cost +80% | $1,201,743 | not within the horizon | -$1,226,833 |
| Year | Usable energy | Revenue | Fixed O&M | Augmentation | Capex / ITC | Net | Cumulative |
|---|---|---|---|---|---|---|---|
| 0 | 100% | -$667,635 | -$667,635 | -$667,635 | |||
| 1 | 100% | $93,570 | -$38,151 | $55,419 | -$612,216 | ||
| 2 | 98% | $79,025 | -$38,151 | $40,875 | -$571,341 | ||
| 3 | 95% | $76,999 | -$38,151 | $38,849 | -$532,492 | ||
| 4 | 92% | $74,973 | -$38,151 | $36,822 | -$495,670 | ||
| 5 | 90% | $72,947 | -$38,151 | $34,796 | -$460,874 | ||
| 6 | 88% | $70,920 | -$38,151 | $32,770 | -$428,104 | ||
| 7 | 85% | $68,894 | -$38,151 | $30,743 | -$397,360 | ||
| 8 | 82% | $66,868 | -$38,151 | $28,717 | -$368,643 | ||
| 9 | 80% | $64,841 | -$38,151 | $26,691 | -$341,952 | ||
| 10 | 100% | $81,052 | -$38,151 | -$105,304 | -$62,403 | -$404,356 | |
| 11 | 98% | $79,025 | -$38,151 | $40,875 | -$363,481 | ||
| 12 | 95% | $76,999 | -$38,151 | $38,849 | -$324,632 | ||
| 13 | 92% | $74,973 | -$38,151 | $36,822 | -$287,810 | ||
| 14 | 90% | $72,947 | -$38,151 | $34,796 | -$253,014 | ||
| 15 | 88% | $70,920 | -$38,151 | $32,770 | -$220,244 |
Year 1 is the first operating year (COD 2029); payback and NPV run from it. What is known: PJM capacity prices are cleared through 2028/29; 2029/30 is still under the price collar and is shown at the 2028/29 price pending the December 2026 auction. What is assumed: everything from 2030/31 on follows labelled scenario paths, not a forecast (capacity $175 / $250 / $500 per MW-day; ELCC sliding to 0.40 by 2035/36 in the base path; arbitrage flat / +5% / +10% per year as an industry-outlook assumption).[34,6,8]
| Year | Delivery year | Capacity $/MW-day | ELCC | Usable | Capacity $ | Arbitrage $ | O&M | Augmentation | Net (base) | Net (low) | Net (high) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2029/30 * | 325 | 0.56 | 100% | $33,387 | $12,585 | -$38,151 | $53,812 | $14,789 | $220,560 | |
| 2 | 2030/31 ** | 250 | 0.54 | 98% | $23,830 | $13,535 | -$38,151 | $31,230 | -$9,074 | $154,279 | |
| 3 | 2031/32 ** | 250 | 0.51 | 95% | $22,045 | $13,847 | -$38,151 | $28,936 | -$9,819 | $129,969 | |
| 4 | 2032/33 ** | 250 | 0.48 | 92% | $20,317 | $14,157 | -$38,151 | $26,697 | -$10,565 | $106,798 | |
| 5 | 2033/34 ** | 250 | 0.45 | 90% | $18,655 | $14,463 | -$38,151 | $24,520 | -$11,310 | $84,766 | |
| 6 | 2034/35 ** | 250 | 0.43 | 88% | $17,051 | $14,764 | -$38,151 | $22,396 | -$12,056 | $82,842 | |
| 7 | 2035/36 ** | 250 | 0.40 | 85% | $15,512 | $15,059 | -$38,151 | $20,333 | -$12,801 | $80,978 | |
| 8 | 2036/37 ** | 250 | 0.40 | 82% | $15,056 | $15,347 | -$38,151 | $19,343 | -$13,547 | $79,174 | |
| 9 | 2037/38 ** | 250 | 0.40 | 80% | $14,600 | $15,626 | -$38,151 | $18,345 | -$14,293 | $77,433 | |
| 10 | 2038/39 ** | 250 | 0.40 | 100% | $18,250 | $20,509 | -$38,151 | -$105,304 | -$71,859 | -$113,633 | $3,518 |
| 11 | 2039/40 ** | 250 | 0.40 | 98% | $17,794 | $20,996 | -$38,151 | $32,656 | -$9,074 | $107,822 | |
| 12 | 2040/41 ** | 250 | 0.40 | 95% | $17,338 | $21,481 | -$38,151 | $31,863 | -$9,819 | $106,946 | |
| 13 | 2041/42 ** | 250 | 0.40 | 92% | $16,881 | $21,962 | -$38,151 | $31,066 | -$10,565 | $106,198 | |
| 14 | 2042/43 ** | 250 | 0.40 | 90% | $16,425 | $22,436 | -$38,151 | $30,264 | -$11,310 | $105,582 | |
| 15 | 2043/44 ** | 250 | 0.40 | 88% | $15,969 | $22,904 | -$38,151 | $29,454 | -$12,056 | $105,104 |
* collar-bounded, auction not yet run. ** scenario path. Capacity = effective MW x usable energy x ELCC x price x 365. Augmentation in year 10.
Energy arbitrage. Each day is solved as a linear program with perfect knowledge of that day's hourly day-ahead prices, carrying the state of charge from the previous day and requiring it to end the day at or above its starting level. That is a ceiling. The achievable row applies a flat 0.80 haircut (range 0.70 to 0.90) to approximate an operator bidding on forecasts. Only the market selected above is used; real-time settlement, intra-hour volatility and co-optimisation with other services are ignored. Prices are hourly zonal LMPs from gridstatus.io (PJM data)[1,23]; real-time zonal values in that feed are estimates[2]. Achievable factor: [12].
Capacity. Revenue = effective MW x ELCC x clearing price x 365, where effective MW = min(power, energy / 4) and the ELCC class is the duration rounded down to 4, 6, 8 or 10 hours. A battery under 4 hours is rated as a fraction of a 4-hour unit (effective MW = energy / 4); this is an assumption to confirm with PJM or a market participant. The low-ELCC row uses 0.40[8]. Clearing prices are the RTO prices at which all four NJ zones cleared; per-LDA overrides are available but empty today. The FERC-approved price collar expires after delivery year 2029/30[6]; PJM's uncapped simulations ran $530 to $555/MW-day, and later prices are treated as unknown. Capacity revenue also depends on obtaining Capacity Interconnection Rights and a Wholesale Market Participation Agreement where applicable; queue timing and network-upgrade costs are not estimated.
What this report is. GridArb is a fast first check, not a bank-ready study. It tells you which sites are worth a $75,000 study deposit and a consultant, and which to drop today. Full method, sources and limits: gridarb /methods. Every rule with the date it was last checked: gridarb /assumptions. Data sources and their terms: gridarb /sources.
Not modelled or not guaranteed. Regulation is shown as its own line only when PJM market results are loaded, with the warning that its prices are expected to fall; other ancillary services are excluded. The NJ GSESP incentive ($60,000 to $81,840 per MW-year for 15 years)[9,10] is a competitive award for projects of at least 5 MW and is shown as a reference range only. The federal ITC (30%, Section 48E, FEOC rules apply)[11] is a cost-side item and is not modelled in detail. Degradation is a flat $5 per MWh discharged, not a cycle-life model. Round-trip efficiency is split equally between charging and discharging.
Generated 2026-10-28 07:02 UTC.