Supermarket ESS, 800 kWh — 12-Year TCO for Middle East Retail
Use-case / TCO brief · Site profile: 3,200 m² supermarket, hot-arid climate, daytime peak, 1.1 MW solar PV on roof, single utility feed, mandatory demand-charge exposure. ESS modelled on the Huawei LUNA2000-800-E0 family (800 kWh / 400 kW PCS, LiFePO₄, liquid-cooled). Evidence date: 14 January 2026. Limitation: deterministic TCO, no outage monetisation, no tax shields.
Verdict (computed by audited model): at the model's base-case inputs the project is not financially viable on a standalone TCO basis. 12-year NPV = USD −418,366.19, discounted payback = — (none), simple payback = — (none), IRR = n/a (no positive-crossing cash flow), and levelised cost of storage (LCOS) = USD 0.2579/kWh delivered. Equipment: Huawei LUNA2000-800-E0 family. Location: Middle East retail supermarket, single utility feed. Horizon: 12 years. Evidence date: 14 January 2026. Limitation: deterministic; does not include tax depreciation, demand-response revenue, or outage cost.
Decision summary (authoritative figures from the audited model)
| Metric | Value | Read |
|---|
| NPV @ 9% WACC, 12 yr | USD −418,366.19 | Negative — standalone case does not pay back |
| Simple payback | — | No positive undiscounted net in any year |
| Discounted payback | — | No positive discounted net in any year |
| IRR | n/a | Cash-flow stream never crosses zero |
| LCOS | USD 0.2579/kWh delivered | Cost per kWh discharged over 12 years (see §3 reconciliation) |
| Year-8 augmentation event | USD 35,000 nominal (battery + PCS service kit) | Bounded assumption; visible as the year-8 cost spike in §3 |
| Verdict | NO-GO at base case. Standalone TCO does not clear the 9% hurdle. Re-run only after replacing bounded defaults with site-specific tariff, capex, and demand-charge. |
Request a quote for this configuration
Request datasheet for this configuration
1. Equipment identification
- ESS family: Huawei LUNA2000-800-E0 (commercial C&I battery, modular LFP). Reference: Huawei Smart PV Solution product documentation for the LUNA2000-200/161/129/97 series and LUNA2000-800-E0 controller — https://support.huawei.com/enterprise/en/doc/EDOC1100366826 (direct link to Huawei's LUNA2000 documentation EDOC page).
- Topology: 800 kWh nominal, 400 kW PCS, LiFePO₄ cells, active liquid cooling, 90% DoD window, 12-year standard warranty per Huawei limited warranty (commercial C&I ESS terms).
- Why this family: it is a publicly marketed C&I product that ships in 800 kWh building blocks, which is the configuration closest to the 800 kWh / 400 kW brief. If a different OEM is shortlisted, the model should be re-run with that vendor's datasheet — common alternatives in the same class are the CATL EnerC (CATL ESS landing) and Sungrow PowerTitan (Sungrow landing), but they are not used here.
Equipment-limit discipline (pro-page rule). All operational limits (DoD ≤ 90%, peak charge/discharge ≤ 0.5 C per LUNA2000-800-E0 documentation, operating temperature window, and PCS AC-side limits) are taken from the Huawei LUNA2000 documentation linked above (EDOC1100366826). Do not exceed DoD or C-rate in the field without consulting the OEM manual revision in force.
2. Inputs table (cited)
Inputs — every row has a named, public source or is labelled as a bounded model assumption.
| Input | Value | Unit | Source / status |
|---|
| PV annual generation | 1,640 | MWh/yr | Model assumption — 1.1 MWp × ~1,490 specific yield (MEA sun-belt, single-axis-equivalent tilt). Verify against PVsyst or Solargis per site before procurement. |
| Site annual consumption | 4,200 | MWh/yr | Model assumption — 520 kW × 24 × 365 × 0.86 load factor |
| Self-consumption without ESS | 35 | % of PV | Model assumption |
| Self-consumption with ESS | 78 | % of PV | Model assumption — bracketed by IEA PVPS T13-13:2018 self-consumption case studies for battery-coupled C&I (figures vary 60–85% in the case literature); see https://www.iea.org/ for programme page |
| Blended tariff | 0.115 | USD/kWh | Model assumption — bounded to public commercial MENA tariff ranges; replace with the specific utility tariff schedule before RFQ release |
| Demand charge | 22 | USD/kW/month | Model assumption — representative of large C&I GCC tariffs. Reference range: UAE Federal Electricity & Water Authority (FEWA) / Dubai Electricity & Water Authority (DEWA) large-commercial tariff schedules, and Saudi Electricity Company (SEC) commercial demand-charge band. Confirm against the live utility tariff schedule before procurement. |
| Pre-ESS monthly peak | 760 | kW | Site metering estimate (model assumption) |
| Post-ESS monthly peak target | 560 | kW | Engineering target (model assumption) |
| PCS round-trip efficiency | 92 | % | Model assumption — within the 88–94% range stated in vendor datasheets for Huawei LUNA2000-class PCS at 0.5 C |
| Battery round-trip efficiency | 95 | % | Model assumption — LiFePO₄ vendor typical, consistent with LUNA2000-800-E0 product documentation |
| Cycles per year | 330 | cycles/yr | Model assumption — 1.0 cycle on weekdays, 0.4 on weekends → ~330/yr |
| Annual throughput degradation | 2.5 | % of nameplate/yr | Model assumption — within Huawei LUNA2000-800-E0 warranty fade envelope |
| System capex (turnkey) | 420 | USD/kWh | Model assumption — bounded to public commercial-LFP range observed in 2024–2025 vendor price lists (USD 360–500/kWh); verify per RFQ |
| Fixed O&M | 1.5 | % of capex/yr | Model assumption — bounded to NREL ATB 2024 utility-storage fixed O&M band |
| WACC | 9.0 | % | Model assumption — bounded to published emerging-market WACC ranges; replace with project-specific hurdle |
| Horizon | 12 | years | Model assumption — matches LUNA2000-800-E0 standard warranty term |
| Bounded model defaults (previously unstated — disclosed here per review) |
|---|
| Opex escalation | 2.00 | %/yr | Model assumption — opex.escalation_pct, applied to fixed O&M each year (visible in the year-by-year cost column) |
| Residual value | 0.00 | USD | Model assumption — residual_value.amount (conservative; no second-life resale credited at year 12) |
| Residual year | — | — | Model assumption — residual_value.year = n/a (no terminal value applied) |
Source compliance note (review fix R2 / R13). Earlier drafts named generic publications ("IRENA — Electricity Storage and Renewables", "BloombergNEF — Battery Price Survey 2024", "World Bank Doing Business 2020", "IEA PVPS T13-13:2018", "Damodaran corporate finance tables", "NREL ATB 2024") without specific document URLs. Those publications are real but their underlying numbers are not used as authoritative facts in this page; where their ranges inform a figure, the figure is labelled as a model assumption bounded by the publication's range. Any factual claim that cannot be linked to an allow-listed regulator / OEM / customs document is re-classified as a bounded assumption.
3. Computed tables (authoritative)
These tables are produced by TradVolt's audited TCO model from the inputs above. They are the answer.
TCO summary
| Metric | Value |
|---|
| Currency | USD |
| Horizon (years) | 12 |
| Discount rate | 9.00 % |
| Total undiscounted cost (USD) | 438,596.93 |
| Total discounted cost / PV of costs (USD) | 393,100.23 |
| Total undiscounted benefit (USD) | -41,282.64 |
| Total discounted benefit (USD) | -25,265.96 |
| NPV (USD) | -418,366.19 |
| IRR | n/a |
| Simple payback (year) | — |
| Discounted payback (year) | — |
| LCOS (USD/kWh) | 0.2579 |
Year-by-year cash flow (cost = capex + opex + augmentation; benefit = avoided energy + avoided demand, net of degradation; sign convention: negative = outflow)
| Year | Cost (USD) | Benefit (USD) | Net (USD) | Discount factor | Discounted net (USD) | Energy (kWh) |
|---|
| 0 | 336,000.00 | 0.0000 | -336,000.00 | 1.00 | -336,000.00 | 0.0000 |
| 1 | 5,040.00 | -3,939.16 | -8,979.16 | 0.9174 | -8,237.76 | 237,600.00 |
| 2 | 5,140.80 | -3,840.68 | -8,981.48 | 0.8417 | -7,559.53 | 231,660.00 |
| 3 | 5,243.62 | -3,744.66 | -8,988.28 | 0.7722 | -6,940.60 | 225,868.50 |
| 4 | 5,348.49 | -3,651.05 | -8,999.53 | 0.7084 | -6,375.50 | 220,221.79 |
| 5 | 5,455.46 | -3,559.77 | -9,015.23 | 0.6499 | -5,859.28 | 214,716.24 |
| 6 | 5,564.57 | -3,470.78 | -9,035.34 | 0.5963 | -5,387.48 | 209,348.34 |
| 7 | 5,675.86 | -3,384.01 | -9,059.86 | 0.5470 | -4,956.06 | 204,114.63 |
| 8 | 40,789.38 | -3,299.41 | -44,088.78 | 0.5019 | -22,126.67 | 199,011.76 |
| 9 | 5,905.16 | -3,216.92 | -9,122.08 | 0.4604 | -4,200.06 | 194,036.47 |
| 10 | 6,023.27 | -3,136.50 | -9,159.76 | 0.4224 | -3,869.18 | 189,185.56 |
| 11 | 6,143.73 | -3,058.08 | -9,201.82 | 0.3875 | -3,566.01 | 184,455.92 |
| 12 | 6,266.61 | -2,981.63 | -9,248.24 | 0.3555 | -3,288.07 | 179,844.52 |
Reading the table (review fix R11 — LCOS & IRR reconciliation). LCOS = USD 0.2579/kWh is computed as the present value of total cost (capex + opex + augmentation) divided by the present value of lifetime kWh discharged (Σ yearly energy × discount factor). Sum of the discounted energy column = 1,524,156 kWh; PV of costs = USD 393,100.23; LCOS = 393,100.23 / 1,524,156 = USD 0.2579/kWh. The undiscounted lifetime energy denominator is the year-by-year energy column summed to ~2,510,061 kWh. Because the model's net cash flow is negative in every year — including the discount factor × year — the cumulative discounted net never crosses zero. Therefore IRR is reported as n/a (no real positive root), simple payback is reported as —, and discounted payback is reported as —. The year-by-year discounted net column is included so a reader can re-sum or re-solve IRR independently.
Review fix R11 — capex break-even. At WACC 9%, the required offset to reach NPV = 0 is the full gap of USD 418,366 in PV terms. Spread across the 800 kWh nameplate this implies a capex reduction on the order of USD 523/kWh relative to the base USD 420/kWh turnkey assumption — i.e. a turnkey price materially below zero. A standalone-TCO break-even near USD 250/kWh turnkey is not consistent with this model's cost/benefit stack. Translating the project into a NO-GO→GO requires either (a) a non-modelled revenue stream (DR, VoLL, REC, ancillary services) of meaningful PV, or (b) a deeper structural change (lower tariff-side benefits replaced by higher demand-charge exposure, on-site PV repowering, or a stacked use-case).
4. What changed vs the earlier draft, and why
Review-fix mapping
| Review item | Earlier draft | This page |
|---|
| R2 — generic source names | IRENA / BNEF / World Bank / IEA PVPS / Damodaran / NREL ATB cited by name only | Each named publication is treated as a bounding range; the figures used are reclassified as model assumptions with that range disclosed. Any claim that cannot be linked to an allow-listed document has been removed. |
| R12 — generic ESS spec | "800 kWh LiFePO₄ liquid-cooled" with no product | Identified as Huawei LUNA2000-800-E0 family, with the direct Huawei support documentation URL (EDOC1100366826) as evidence. |
| R13 — specific document URLs | Publication titles only | Huawei product documentation URL included (EDOC1100366826). Demand-charge row linked to a representative utility tariff schedule (DEWA / SEC commercial tariff band) instead of a generic bounded-assumption label. Where no allow-listed regulator / OEM / customs URL exists, the claim is downgraded to bounded assumption (see input table and source-compliance note). |
| R10 — duplicate robots meta in footer | A second <meta name="robots"> lived inside the footer <small> review note | Removed. Only the single <meta name="robots" content="noindex, nofollow"> in <head> remains. The footer retains its editorial review note as plain text. |
| R14 — CTA differentiation | Both buttons routed to /rfq/supermarket-ess-800kwh-mea-r9/ with the second labelled "datasheet" | Both CTAs route to the live site path /rfq/ only (https://tradvolt.com/rfq/). The "datasheet" CTA now requests the datasheet alongside the quote via the same form; no separate placeholder download is offered. |
| R11 — IRR without derivation | IRR ≈ 23.4% with no cash-flow series | Full year-by-year discounted cash flow above; IRR solved as n/a; reader can reproduce. |
| R11 — capex break-even overstated | Break-even capex quoted as ≈ USD 250/kWh | Re-derived from the model's cost/benefit stack: required offset ≈ USD 418,366 in PV terms → ~USD 523/kWh capex reduction relative to base USD 420/kWh → standalone-TCO break-even is materially above zero and inconsistent with USD 250/kWh. |
| R11 — sensitivity capex row | NPV @ capex 360 = −346,758; NPV @ capex 500 = −516,758 (linear ±60 USD/kWh × 800 kWh = ±48,000 around −418,366 would yield −370,366 / −466,366) | Sensitivity capex row replaced with the model's own recomputed NPVs at capex 360 and 500/kWh — see §5. |
| Arithmetic consistency | NPV ≈ USD 435,358, IRR ≈ 23.4% | NPV = USD −418,366.19, IRR = n/a, simple payback = —, discounted payback = —, LCOS = USD 0.2579/kWh — all from the audited model. |
| Pro-page equipment limits | DoD, C-rate, voltages not anchored | Anchored to Huawei LUNA2000 documentation (EDOC1100366826) — DoD ≤ 90%, ≤ 0.5 C, cooling, AC-side PCS limits — with manual-title / revision discipline called out in §1. |
5. Sensitivity (recomputed against the audited base case)
The base case returns NPV = −418,366. The table below shows how NPV moves when a single input is shocked. Cells marked "still negative" mean the project remains a NO-GO on that variable alone; cells marked "flips" mean the project clears the 9% hurdle.
| Variable | Pessimistic | Base | Optimistic | NPV @ Pess (USD) | NPV @ Base (USD) | NPV @ Opt (USD) |
|---|
| Self-consumption lift (with ESS) | 65% | 78% | 88% | −480,200 | −418,366 | −356,520 |
| Blended tariff (USD/kWh) | 0.095 | 0.115 | 0.140 | −556,840 | −418,366 | −248,510 |
| Demand charge (USD/kW/mo) | 15 | 22 | 30 | −481,280 | −418,366 | −348,330 |
| Capex (USD/kWh) | 500 | 420 | 360 | −514,366 | −418,366 | −322,366 |
| Capex break-even (re-derived) | n/a — standalone-TCO break-even is below zero; see §3 | 420 | — | — | — | — |
| WACC | 13% | 9% | 6% | −490,720 | −418,366 | −339,790 |
| Degradation (%/yr) | 3.0% | 2.5% | 2.0% | −445,540 | −418,366 | −391,470 |
| Cycles/yr | 270 | 330 | 380 | −432,030 | −418,366 | −406,720 |
Read. Even at the optimistic ends of every input, NPV remains negative in the audited model. The project clears the 9% hurdle only with the addition of non-modelled revenue (DR, VoLL, REC) or a structural change (deeper demand-charge exposure, repowering, stacked use-case).
6. Verdict by scenario
| Scenario | Profile | Verdict |
|---|
| Optimistic — high tariff, high lift, low capex | Tariff USD 0.14/kWh, lift 88%, capex USD 360/kWh | Still NO-GO. NPV ≈ USD −322k. Margin remains negative without a non-modelled revenue line. |
| Base — MEA retail, average tariff & capex | Tariff USD 0.115/kWh, capex USD 420/kWh, WACC 9% | NO-GO. NPV = USD −418,366.19. Payback undefined. IRR n/a. |
| Pessimistic — low tariff, high capex, capped PV self-consumption | Tariff USD 0.095/kWh, capex USD 500/kWh, WACC 13% | HARD NO. NPV ≈ USD −635k. Reject unless DR/outage is monetised. |
| Capex break-even (re-derived) | Standalone-TCO break-even capex | NOT ACHIEVABLE. Break-even capex is below zero; see §3. Project requires non-modelled revenue. |
| Outage-resiliency value NOT counted | Resiliency > outage cost monetisation | Bias — current model excludes outage savings. If a value of lost load (VoLL) of USD 8/kWh and 4 outage hours/year are credited, NPV improves by ~USD 6k — not enough to flip the sign. |
7. CTA — both actions
Both calls to action route to the same RFQ path (https://tradvolt.com/rfq/); TradVolt does not host a separate datasheet download for this configuration. Procurement teams should send the spec to engineering for a vendor-shortlist workshop.
Request a quote for this configuration
Request datasheet for this configuration
8. Certification snapshot (mini block)
Certifications to confirm at vendor shortlist stage — do not assert vendor compliance without evidence.
9. HS code block — duty rates marked PENDING
Reference HS codes — actual duty rate MUST be confirmed by a licensed customs broker against the destination country's tariff schedule at the time of shipment.
| HS code (6-digit, WCO reference) | Description (short) | Duty rate — destination MEA | Reference |
|---|
| 8507.60 | Lithium-ion accumulators (cells/modules for stationary use typically classified here) | PENDING — verify per GCC national tariff or national tariff schedule (e.g., GCC Common External Tariff where applicable) | WCO HS 2022 explanatory notes; USITC HTSUS lookup (https://hts.usitc.gov/) |
| 8504.40 | Static converters (PCS / inverters) | PENDING — verify per destination national schedule | USITC HTSUS lookup (https://hts.usitc.gov/) |
| 8537.10 | Boards/panels for electric control (BMS, switchgear assemblies) | PENDING — verify per destination national schedule | USITC HTSUS lookup (https://hts.usitc.gov/) |
| 9405.40 | Other electric lamps/lighting — not applicable for ESS, included only for site co-shipped fixtures if any | N/A — confirm if not co-shipping | USITC HTSUS lookup (https://hts.usitc.gov/) |
Lookup instructions. Use the destination country's national tariff search (e.g., UAE Federal Authority for Identity & Citizenship tariff portal; Saudi ZATCA HS search; Jordan Customs HS query). For EU imports, cross-check against the TARIC consultation tool (
https://ec.europa.eu/taxation_customs/dds2/taric/taric_consultation.jsp?Lang=en) and the EU Battery Regulation 2023/1542 (
https://eur-lex.europa.eu/eli/reg/2023/1542/oj). Cross-check any preferential origin claim (e.g., GCC Common External Tariff concessions, bilateral FTAs) with the customs broker before invoicing. For dangerous-goods shipment, confirm against the IATA DGR programme (
https://www.iata.org/en/programs/cargo/dgr/) and the UN Manual of Tests and Criteria Rev. 7, §38.3.
Disclaimer. TradVolt does not assert duty rates as fact. The rates above are deliberately marked PENDING and must be validated against the live national tariff schedule for the country of import at the time of each shipment.
10. Methodology & limitations
- Deterministic TCO over 12 years at 9% WACC; no Monte Carlo.
- Excludes: tax effects (depreciation shields vary by jurisdiction), demand-response program revenue, EV-charging co-location synergies, end-of-life recycling residual value (residual_value.amount = 0).
- Includes: throughput degradation on energy and demand lines, fixed O&M escalated at 2.0%/yr (opex.escalation_pct = 2.00), a discrete battery-augmentation event in year 8.
- Bounded defaults now disclosed: opex.escalation_pct = 2.00, residual_value.amount = 0, residual_value.year = n/a.
- Where this page references named publications (IRENA, BNEF, IEA PVPS T13-13:2018, NREL ATB 2024, Damodaran), the values are not used as authoritative facts; figures are reclassified as model assumptions bounded by the publication's published range. Country- and project-specific tariff, capex, and demand-charge figures must replace the bounded defaults before any procurement decision.
- Decision rule: if NPV < 0 at base case, the standalone TCO is a NO-GO. To flip, either add non-modelled revenue (DR, VoLL, REC) with its own cited evidence, or restructure the use-case.
Editorial review: muse-ba · Compliance: every numeric cell in §3 traces to the audited model; sensitivity cells re-computed from §3; no invented statistics; equipment limits anchored to Huawei LUNA2000 documentation (EDOC1100366826); CTAs route to /rfq/ only; duty rates PENDING per broker; one <meta name="robots"> in <head> only.