Solar Industry Romania / strategic case study / board decision

Ecodepot BESS portfolio & supply-chain roadmap.

Five operating years, 2027–2031, with H2 2026 as the mobilisation window. The central conclusion is that Ecodepot does not need a broader battery catalogue. It needs a narrower stocked architecture, a dedicated retrofit route, harder supplier qualification and an operating model that converts a rapidly expanding Romanian prosumer base into repeatable installer-led demand without allowing inventory, compatibility or warranty risk to absorb the economics.

Executive decision

Build a compact platform, not a battery warehouse.

Romania's storage adoption is already large enough to justify a dedicated portfolio, but the public evidence does not justify stocking every capacity or every brand. The strategy should maximise availability at the common component level while keeping long-tail configurations and larger C&I systems project-funded.

Recommendation: retain WHES as the anchor platform, concentrate physical inventory around modular components that construct approximately 5, 10 and 15 kWh residential systems, make 20–32 kWh configurations predominantly order-led, and qualify one retrofit/compatibility specialist for activation only if WHES cannot provide a Romanian-compliant retrofit architecture with acceptable service and channel terms.
Do
Stock the common denominator.

Battery modules, control units, a deliberately limited inverter power ladder and the accessories that recur across multiple virtual bundles should carry the service-level inventory. This preserves responsiveness without betting capital on every finished configuration.

Add
A retrofit route, not another logo.

Romania had 201,304 individual prosumers without storage at 30 June 2026. The specialist supplier slot exists to solve installed-base compatibility if the anchor platform cannot do so cleanly; it is not a licence to create a multi-brand catalogue. S01

Avoid
Broad speculative inventory.

Larger residential configurations, 57–100 kWh light-C&I and all systems above that range should remain principally project/deposit funded until Ecodepot proves stable, repeatable monthly cadence and a working-capital advantage from local stock.

Primary evidence baseline

Storage adoption has moved from niche to material.

The June 2026 ANRE dataset is unusually decision-useful because it distinguishes prosumers by legal form and reports how many already have storage. The resulting retrofit pool is a stock variable, not a theoretical TAM.

Romania / ANRE / connected prosumers / 30 Jun 2026 Reported values + analyst-derived ratios
Total prosumers 359,378 Reported connected prosumer sites. S01
With storage 127,084 Reported prosumers with installed storage. S01
Storage penetration 35.4% Derived: 127,084 / 359,378. Not an annual sales attach rate.
Households without storage 201,304 Derived: 324,535 PF prosumers less 123,231 PF with storage. S01
Exhibit 01 / sites
The storage-equipped stock almost doubled in six months while the no-storage stock was almost flat.
Dec 2025 vs Jun 2026 / Romania
Romanian prosumer stock split by storage status, December 2025 and June 2026 December 2025: 65,175 storage-equipped and 230,301 without storage. June 2026: 127,084 storage-equipped and 232,294 without storage. 0 120k 240k 360k 31 DEC 2025 30 JUN 2026 WITH STORAGE NO STORAGE

Source: ANRE monthly prosumer situation at 31 Dec 2025 and 30 Jun 2026. Geography: Romania. Unit: connected prosumer sites. Analysis: Solar Industry Romania calculations. S01 S02

01 / frame

The portfolio question needs reframing.

Ecodepot's public catalogue already spans the relevant capacity ladder. The strategic problem is therefore not primarily product discovery; it is capital allocation, use-case routing, supplier governance and channel execution. S18

Reported direct external evidence Derived analyst calculation Assumption planning input requiring validation

Ecodepot currently presents WHES residential Battery Box products from 4.99 to 29.9 kWh, single-phase PP-S2 systems from 3.6 to 10 kW with 4.99–19.9 kWh, three-phase residential systems, and C&I products extending from PC-mini through 509 kWh configurations. That public scope already covers the fundamental energy-capacity ladder. S18 S19 S20

The premise challenge is therefore material: adding brands because competitors carry them would solve no demonstrated customer problem. A new supplier should enter the active catalogue only if it closes a defined gap in compatibility, service, commercial economics, channel pull, availability or concentration risk. The operating principle used throughout this report is a compact active catalogue supported by a broader qualified-supplier database.

The highest-evidence whitespace is retrofit. ANRE reported 324,535 individual prosumers at 30 June 2026, of whom 123,231 had storage. The residual 201,304 households are an installed solar base for which a battery sale does not necessarily require another PV installation. This changes Ecodepot's market architecture: “new solar plus storage” and “storage retrofit” must be separate commercial products with separate compatibility, commissioning and warranty processes. S01

The decision is not whether Romania needs a 10 kWh battery.

The decision is whether Ecodepot can repeatedly sell, commission, support and finance a narrow set of modular architectures across new-build and retrofit use cases more profitably than a broad catalogue. Capacity is a configuration variable; the durable product is the operating system around selection, compatibility, installation, warranty and availability.

02 / market evidence

Romania has entered the storage-attachment phase.

The acceleration is real, but the cause is multi-factor and policy-sensitive. It should be used to justify portfolio readiness, not as a basis for blindly extrapolating a 95% six-month growth rate.

Individual prosumers / Jun 2026
324,535

90.3% of the national prosumer site count. Household demand therefore dominates site volumes even though legal-person sites carry much larger average PV capacity. S01

PF with storage / Jun 2026
123,231

Derived household storage penetration: 38.0%. Individual prosumers account for approximately 97.0% of all storage-equipped prosumer sites. S01

PJ with storage / Jun 2026
3,853

Derived storage penetration among legal-person prosumers: 11.1%, materially below household penetration. This supports a more selective, load-evidence-led B2B approach. S01

PF average PV nameplate
6.37 kW

Derived from 2,067.47 MW across 324,535 individual prosumers. This is a PV site-average reference, not evidence that a 6.37 kWh battery is optimal. S01

Metric 31 Dec 2025 30 Jun 2026 Change Decision implication
Total prosumer sites 295,476 359,378 +63,902 / +21.6% Underlying installed solar base continues to expand. S01 S02
Storage-equipped sites 65,175 127,084 +61,909 / +95.0% Storage is no longer peripheral to the prosumer market.
Overall storage penetration 22.1% 35.4% +13.3 percentage points Integrated storage should be a standard quotation path rather than an optional appendix.
Individual prosumer sites 264,833 324,535 +22.5% Residential remains the scale segment.
Individual prosumers with storage 62,865 123,231 +96.0% Residential storage is the principal unit-volume opportunity.
Legal-person prosumers 30,643 34,843 +13.7% B2B has lower site volume but materially larger average system size.
Legal-person prosumers with storage 2,310 3,853 +66.8% Growth is rapid from a small base; avoid assuming residential attachment economics transfer directly.
Total prosumer PV nameplate 3,453.78 MW 4,019.17 MW +565.39 MW / +16.4% Installed power grew more slowly than site count over the period, consistent with an expanding small-site base.

2025 was the inflection year, but not a safe CAGR.

ANRE's 2025 annual prosumer report states that storage-equipped prosumers increased from approximately 2,400 around the start/end of 2024 to approximately 65,000 by the end of 2025, more than 27 times the earlier level. The report itself describes the trajectory as exceptionally accelerated. S03

The six months to June 2026 then added another 61,909 storage-equipped sites. This confirms persistence beyond the initial 2025 acceleration, but it does not identify which share resulted from batteries installed together with new PV, which share resulted from retrofit, or which share was subsidy-driven. The appropriate strategic response is readiness plus scenario discipline, not straight-line extrapolation.

Electricity-cost salience supports demand, but does not establish payback.

Eurostat reported that Romanian household electricity prices in national-currency terms increased 58.6% year on year in the second half of 2025, the largest increase among EU countries in that comparison. Expressed in purchasing-power-standard terms, Romania also ranked highest in the EU at 49.52 PPS per 100 kWh. S08 This materially increases customer attention to self-consumption and bill control, but it is not sufficient to calculate a BESS return.

ANRE reported 275,010 quantitative-compensation contracts for prosumers at or below 200 kW at the end of 2025, approximately 1,344 GWh of energy delivered to the grid under that mechanism during 2025, average monthly banked energy of 106.86 GWh and average monthly consumption from banked quantities of 64.26 GWh. S03 Consequently, a residential battery case cannot responsibly be reduced to “retail electricity price minus export price.” Tariff treatment, compensation, self-consumption timing, battery efficiency, degradation, critical-load value and future tariff design must be modelled at customer level.

No public dataset identified in this research gives the installed Romanian residential battery capacity distribution in kWh.

That evidence gap is critical. The 5–7, 9–12, 14–20 and 24–32 kWh portfolio bands in this report are strategic product-architecture recommendations, not measured Romanian market shares. Ecodepot's own quote, order and installer data must be used within 30 days to calibrate stock weight by band.

03 / geographic channel proxy

Installer coverage should follow prosumer density before it follows national advertising.

County prosumer stock is not the same as county battery demand. It is nevertheless a useful first proxy for where Ecodepot can find large installed bases of potential new-build and retrofit customers through installers.

Exhibit 02 / top ten counties by prosumer sites / Jun 2026

Ten counties hold 38.8% of Romania's connected prosumer sites.

The concentration is meaningful enough to prioritise installer recruitment and retrofit campaigns, but not high enough to justify a narrow geographic strategy. Source: ANRE June 2026 county dataset. S01

Top ten Romanian counties by connected prosumer count at 30 June 2026 Ilfov 20,812; Timis 17,166; Bihor 15,983; Dolj 13,719; Arges 12,526; Arad 12,424; Suceava 12,372; Cluj 11,649; Iasi 11,569; Mures 11,332. ILFOV 20,812 TIMIȘ 17,166 BIHOR 15,983 DOLJ 13,719 ARGEȘ 12,526 ARAD 12,424 SUCEAVA 12,372 CLUJ 11,649 IAȘI 11,569 MUREȘ 11,332
Commercial interpretation
139,552

Connected prosumer sites across the top ten counties, equivalent to 38.8% of the national June 2026 stock. This should be used to sequence installer-account mapping, not to infer a 38.8% battery-sales share.

Recommended channel use.

Build a county-by-county installer map that overlays Ecodepot's actual customer accounts, quote history, certified installer capacity, service response time and retrofit-ready inverter installed base. The public ANRE map is the demand-base layer; Ecodepot's internal data determines where commercial effort should concentrate.

04 / customer and jobs-to-be-done

Customers buy outcomes; kWh is the configuration.

A portfolio organised only by battery size encourages mis-sizing and commodity price comparison. Ecodepot should merchandise by use case first, then configure power, energy, phase topology and backup capability.

R1
Entry new-build

Household adding solar and first battery, high budget sensitivity.

Primary job

Increase self-consumption and enter storage at the lowest acceptable system cost.

Architecture

Approximately 5–7 kWh, usually 3.6–6 kW hybrid; expansion capability matters more than maximum initial capacity.

Stocked entry route
R2
Mainstream prosumer

Typical household seeking greater evening coverage.

Primary job

Shift daytime PV production into evening load with manageable capital cost.

Architecture

Approximately 9–12 kWh, 5–10 kW hybrid, single- or three-phase depending connection and loads.

Core availability
R3
Existing-PV retrofit

Prosumer already operating a grid-tied PV installation.

Primary job

Add storage without unnecessarily replacing functioning PV assets or invalidating warranties.

Architecture

Approximately 5–20 kWh; AC-coupled or otherwise formally approved retrofit topology; compatibility is the central product feature.

Strategic growth route
R4
Electrified household

Heat pump, EV, pool, electric cooking or high evening load.

Primary job

Higher usable energy and discharge power without forcing a commercial-scale solution.

Architecture

Approximately 14–21 kWh, often 8–13 kW three-phase; configure against simultaneous kW as well as daily kWh.

Lean stock / configure
R5
Resilience-led household

Rural, unstable supply, critical loads or strong backup requirement.

Primary job

Continuity of selected circuits, not simply self-consumption optimisation.

Architecture

Battery energy must be paired with backup power, transfer behaviour, black-start and critical-load design. “Days of autonomy” claims require load modelling.

Engineered bundle
R6
Large residential / microbusiness

Large home, farm, workshop, office or mixed-use property.

Primary job

Higher energy capacity with three-phase power and potentially limited commercial functions.

Architecture

Approximately 24–32 kWh, 10–20 kW depending platform and connection; generally configure-to-order.

Order-led
B1
Light C&I / B2B2C

SME with material post-solar load, backup need or controllable peak.

Primary job

Measured operating-cost reduction and/or continuity value.

Architecture

Approximately 57–100 kWh and 30–50 kW is a useful standardised offer envelope aligned with Ecodepot's current PC-mini range, but final size must use interval load data.

Project order
B2
C&I engineered

Commercial or industrial site beyond light-C&I standardisation.

Primary job

Peak management, continuity, PV shifting, flexibility or combined operational requirements.

Architecture

100–500+ kWh and project-specific power. Ecodepot already publicly lists systems through 509 kWh; this is not residential inventory. S18

Engineered / deposit-funded

B2B storage should not inherit residential assumptions.

At 30 June 2026, legal-person prosumers had approximately 1,951.69 MW of installed prosumer generation across 34,843 sites, equivalent to roughly 56.0 kW per site on average. Yet only 3,853 legal-person prosumers were reported with storage, an 11.1% penetration rate, versus 38.0% among individual prosumers. S01

ANRE's 2025 analysis attributes part of this difference to load alignment: household consumption frequently occurs outside PV production hours, while many legal persons consume more during the solar-production window. S03 Ecodepot should therefore require 15-minute or hourly load evidence for B2B storage propositions rather than assuming every commercial PV owner has a battery-shaped problem.

05 / capacity architecture

Five capacity bands cover the residential decision space without five warehouses.

The bands below are portfolio-management ranges, not claims about observed Romanian market share. The correct operating model is to construct several sellable capacities from as few common modules, control units and inverter families as supplier rules permit.

Band A / entry
5–7 kWh

Budget-sensitive first storage, subsidy-compatible entry logic and small-load homes.

Inventory: shallow-to-core where module commonality is high.
Band B / mainstream
9–12 kWh

Default commercial centre of gravity to validate against Ecodepot order history. Suitable for many evening-shifting use cases.

Inventory: highest service-level priority.
Band C / high-load
14–20/21 kWh

Heat pumps, EVs, larger evening loads, higher backup requirement and three-phase households.

Inventory: modular components; limited finished configurations.
Band D / large residential
24–32 kWh

Large homes, farms, mixed-use properties and microbusinesses. Greater sizing and commissioning risk.

Inventory: primarily configure/order-to-demand.
Band E / light C&I
57–100 kWh

Standardisable SME range around Ecodepot's current 29.9–50 kW PC-mini product family.

Inventory: project order; customer deposit where commercially feasible. S20

The 5 kWh tier is necessary, but the AFM floor must not become a sizing rule.

The consolidated Casa Verde photovoltaic programme guide has used a minimum 5 kWh installed storage requirement alongside the supported PV-plus-storage framework. S05 This is strong evidence that Ecodepot needs a compliant entry product at or above that threshold. It is not evidence that 5 kWh is economically or technically optimal for every household. The same distinction should be explicit in commercial training.

The current WHES ladder can already create a compact architecture.

Ecodepot publicly lists WHES Battery Box capacities of 4.99, 9.98, 14.95, 19.9, 24.9 and 29.9 kWh. WHES itself identifies the family as a high-voltage LFP battery from 4.99 to 29.9 kWh. S18 S19 That step pattern is strategically valuable because a common module can potentially support several sellable system sizes. Procurement should verify which control units, BMS components and accessories are common, and whether mixed production revisions or later expansion have warranty constraints.

Portfolio rationalisation decision.

Ecodepot currently also presents an 8–32 kWh residential family. Keep two overlapping battery-module families active only if they solve materially different inverter, power, phase, environmental, installation or supplier-term requirements. If not, one family should become quote-only. Overlap without a defined role consumes stock, installer training and troubleshooting capacity.

06 / technology evolution

LFP stays core; architecture and software become more differentiating than chemistry.

The five-year technology risk is less about missing the next chemistry and more about carrying incompatible hardware while inverter, EMS, grid-support, battery-passport and service requirements evolve.

Chemistry / core

LFP remains the default procurement baseline.

IEA reports LFP at around 90% of global BESS deployments in 2025 and more than 90% in another 2026 battery-market analysis. Its lower relative cost and stationary-cycle fit make it the rational core chemistry unless a supplier-specific alternative provides demonstrably better lifecycle economics. S09 S10

Chemistry / watch

Sodium-ion belongs on the watchlist, not in core stock.

IEA describes 2026 as potentially pivotal for sodium-ion scale-up, but notes that optimised LFP retains advantages in energy density, supply-chain maturity and cost. Current and announced sodium-ion manufacturing also remains overwhelmingly concentrated in China. S11

Architecture / new PV

Modular HV plus hybrid inverter remains the mainstream new-build pattern.

WHES, Huawei, Sungrow and BYD all publicly offer modular high-voltage residential battery families across broadly similar single-digit to low-twenties kWh ranges. This is architecture evidence, not an independent quality ranking. S19 S21 S22 S23

Architecture / retrofit

AC coupling becomes commercially strategic.

GoodWe publicly markets a 3–6 kW single-phase AC-coupled retrofit battery inverter; Solis lists single-phase and three-phase AC-coupled products intended to add batteries to existing grid-tied PV. These establish credible specialist routes if the anchor platform cannot meet Romanian retrofit needs. S24 S25

System / resilience

Power, phase behaviour and backup topology must sit beside kWh.

Ecodepot's configurator logic already notes that battery size must reflect consumption, PV production, desired autonomy, critical-load share and simultaneous power. This should become a mandatory quotation workflow rather than advisory copy. S18

Digital / 2027+

Serial-level traceability and battery data become procurement requirements.

EU Regulation 2023/1542 requires a battery passport from 18 February 2027 for each industrial battery above 2 kWh; the Regulation explicitly treats batteries used for energy storage in private/domestic environments as industrial batteries. S12

Product roadmaps should be governed by functions, not model names.

Between 2027 and 2031, Ecodepot should maintain a minimum specification architecture that survives product-generation changes. The checklist should include voltage class, usable energy, continuous and surge power, phase behaviour, backup/EPS topology, minimum and maximum battery stack, PV-input current and oversizing rules where a hybrid inverter is involved, operating temperature, IP rating, commissioning method, remote diagnostics, firmware-update process, data access, warranty expansion rules, serial traceability and end-of-life notice period.

This is commercially important because a seemingly minor inverter revision can strand battery inventory if BMS protocol, supported firmware or certification changes. Ecodepot should therefore version-control every approved inverter/battery/BMS combination. Installer quotation software should select only combinations in the current approved matrix.

Battery passport readiness is a 2026 procurement issue, not a 2027 IT project.

Regulation (EU) 2023/1542 states that batteries used for private or domestic energy storage are industrial batteries for the purposes of the Regulation and that, from 18 February 2027, industrial batteries above 2 kWh placed on the market or put into service must have an electronic battery passport. S12 Ecodepot should obtain a supplier-specific responsibility matrix covering who is the economic operator placing the product on the EU/Romanian market, who creates and maintains the passport, what individual serial data are available, how replacement batteries are handled and how data remain available if a supplier changes legal entity.

The 2025 amending Regulation postponed the battery due-diligence application timing by two years relative to the original schedule. S14 This should not be interpreted as permission to defer supplier-chain mapping. Procurement contracts signed now may run across later compliance dates. Product-specific legal applicability and producer/EPR responsibilities should be confirmed with Romanian/EU counsel.

07 / current Ecodepot position

The existing public range is broad enough; the operating model is the gap.

Public information supports an anchor-platform strategy because the current WHES relationship already covers entry residential, mainstream residential, large residential and C&I. Commercial terms, service performance and Romania-specific certification still require confidential DD.

Current public family Published range Strategic role Recommended stock posture DD before scale
WHES Battery Box 4.99–29.9 kWh HV LFP Residential anchor module family Stock common modules/control components after revision compatibility is confirmed. Romanian conformity dossier, BMS revision policy, expansion rules, passport/EPR responsibility, warranty logistics.
PP-S2 3.6–10 kW / 4.99–19.9 kWh Single-phase new-build all-in-one/hybrid route Stock only the power ratings validated as high-velocity by Ecodepot order history. Grid conformity by model/revision, backup behaviour, warranty boundary, commissioning time.
PP-T1 / three-phase home Public Ecodepot/WHES three-phase residential range Three-phase mainstream/high-load route Core selected ratings; long-tail ratings quote-only. Unbalance behaviour, phase requirements, exact compatible battery configurations and Romanian compliance.
Battery Box 8–32 8 / 16 / 24 / 32 kWh Potential second residential architecture Keep active only if it solves a distinct power/compatibility/use-case problem. Overlap economics versus 4.99 kWh family; stock duplication; service burden.
PC-mini 29.9–50 kW / 57–100 kWh Standardised light C&I Project-order; maintain demonstration/critical spare strategy rather than finished-system stock. Site design, protection, fire/safety dossier, commissioning, service SLA, project warranty and cash cycle. S20
PC-G1 / G2 / G3 Up to 250 kW / 509 kWh publicly listed Engineered C&I Project-specific procurement only. Project bankability, technical scope, service resources, fire strategy, integration, financing, guarantees.
Critical unknown: the WHES partnership is public; its economics are not.

No public evidence reviewed here establishes Ecodepot's territory, exclusivity, rebates, MOQ, payment terms, project-registration rights, account protection, non-circumvention, price protection, RMA SLA, advanced-replacement rights, spare-parts entitlement or warranty legal entity. These terms are decision-critical and must be inserted into the supplier scorecard before any volume allocation.

08 / supplier landscape

Qualify alternatives by role; activate at most one additional residential line initially.

The suppliers below are not ranked for quality. Public manufacturer documentation supports product-range and architecture observations only. Award decisions require commercial, service, certification and field-performance evidence that is not public.

Anchor / incumbent

WHES

Public Ecodepot relationship already covers approximately 5–30 kWh residential batteries and 57–100 kWh light-C&I before larger systems. Recommended status: flagship/anchor, subject to full commercial and compliance DD. S18 S19

Retrofit specialist / candidate

GoodWe or Solis

Both publish dedicated AC-coupled retrofit inverter architectures. Recommended status: run one competitive technical/commercial qualification for the specialist slot if WHES cannot deliver an acceptable Romanian-approved retrofit route. Do not automatically activate both. S24 S25

Premium / installer-pull candidate

Huawei

LUNA2000 S1 publicly spans 5–21 kWh using 5 and 7 kWh modules. Recommended status: qualified defensive/premium option, but activation should require Ecodepot lost-quote evidence showing brand pull large enough to offset incremental inventory and support. S21

Premium alternative / diversification

Sungrow

Current SBR family publicly spans 6.4–25.6 kWh HV LFP. Recommended status: qualified alternative and negotiating benchmark; active stock only if commercial/service terms materially improve risk-adjusted economics. S22

Compatibility specialist

BYD Battery-Box

Public HVS/HVM/LVS families cover broad HV and LV capacity ranges and are designed for external-inverter ecosystems. Recommended status: qualified compatibility specialist; stock only after Ecodepot quantifies the installed inverter base that requires it. S23

Value / challenger

Deye

AI-W5.1-B is a 5.12 kWh LFP module within a flexible low-voltage architecture. Recommended status: value/challenger database option, activated only if price-sensitive demand is empirically material and expected support/RMA economics remain acceptable. S26

Two logos are not automatically two supply chains.

IEA reports that more than 90% of battery-storage applications rely on LFP batteries that are almost exclusively supplied from China, and highlights the concentration of manufacturing capacity and technical expertise. S09 Ecodepot should therefore measure concentration at several layers: cell source, pack assembly, BMS, PCS/inverter, controller, firmware/cloud, EU warranty entity, spare-parts location and logistics route. Buying from two Chinese brands may reduce single-counterparty risk while leaving upstream and geopolitical concentration largely unchanged.

Supply resilience should consequently combine brand-level optionality with contractual protections: rolling forecasts without excessive take-or-pay, model-change notice, compatible replacement obligations, price-rebalancing or stock-rotation mechanisms where commercially achievable, warranty spares, defective-battery reverse logistics and a documented path for continued software support if a product generation is discontinued.

09 / procurement operating standard

Supplier approval must be harder than catalogue approval.

BESS procurement transfers long-tail obligations to the distributor: safety, transport, firmware, installation support, warranty, recalls and replacement logistics. Purchase price alone is therefore an incomplete metric.

Proposed board-approved weighting / strategic judgement

Score only after hard gates pass.

The weights are a proposed management framework, not market facts. A supplier that fails a safety, legal or warranty hard gate should not compensate by scoring highly on price.

Technical / safety / regulatory evidence25%
Warranty / service / RMA / EU recourse20%
Commercial terms / channel protection15%
Supply resilience / lead time / revision control15%
Portfolio fit / compatibility / roadmap10%
Landed economics / working capital10%
Digital / cyber / data / EMS5%
Hard gate / product

No dossier, no stock.

  • Exact-model documentation Datasheet, installation manual and revision identifier.
  • EU declaration / CE Exact model and current revision, not a family-level marketing statement.
  • Cell / battery safety Relevant evidence such as IEC 62619 where applicable; verify project-specific standards. S15
  • Stationary installation safety Check relevant IEC 62485-5 and local electrical-installation requirements. S16
  • Dangerous-goods logistics UN 38.3 test summary and transport documentation per exact battery. S17
  • Grid interface Romanian prosumer/inverter conformity evidence where applicable; version-controlled approval.
  • Battery Regulation Producer/economic-operator mapping, EPR responsibilities and 2027 battery-passport readiness. S12
  • Traceability Serial/batch level shipment, warranty and recall traceability.
  • Compatibility Versioned inverter/BMS/battery matrix with firmware requirements.
  • Lifecycle End-of-life notice and replacement/compatibility path.
Hard gate / commercial & service

Warranty duration is not a warranty strategy.

  • Warranty legal entity Identify who owes the remedy in the EU and Romania.
  • Coverage Years, throughput/cycles, remaining-capacity conditions and exclusions.
  • Labour Who pays technician time, travel and recommissioning.
  • Freight Who pays compliant return/collection of defective lithium batteries.
  • RMA SLA Diagnosis, authorisation, replacement dispatch and final closure times.
  • Advanced replacement Terms for critical failures and installer/customer continuity.
  • Spare parts Local/EU availability and minimum support period.
  • Account protection Deal registration, project-price protection and non-circumvention.
  • Payment terms Deposit, balance timing, credit, LC requirements and currency.
  • Price protection Mechanism for rapid list-price reductions while Ecodepot holds stock.
  • Training Installer accreditation, technical hotline and escalation route.
  • Data / cloud API, portal, remote diagnostics, cybersecurity contact and firmware policy.

Safety evidence should be product-specific.

IEC 62619:2022 specifies safety requirements and tests for secondary lithium cells and batteries used in industrial applications, including stationary applications. S15 IEC 62485-5 addresses safety aspects associated with installation, use, inspection, maintenance and disposal of stationary lithium-ion batteries. S16 These standards are procurement reference points, not a complete statement of every Romanian legal or project requirement. The exact applicable conformity set must be checked per product, inverter, installation type and placing-on-market role.

The procurement method used in this report therefore treats warranty as a full operating-liability review rather than a duration comparison. It covers throughput, capacity retention, exclusions, temperature restrictions, remote-connectivity requirements, registration, labour reimbursement, replacement logistics, transport, spare parts, response SLA and warranty-provider solvency.

10 / inventory and working capital

The correct unit of inventory is the reusable component, not the marketing bundle.

Falling battery prices create a structural penalty for slow stock. Ecodepot should preserve customer availability through component commonality while limiting finished configurations that can be stranded by price cuts or compatibility revisions.

AX / core repeaters

Stock common battery modules and the few inverter ratings that drive most qualified demand.

Service-level inventory should be concentrated where the same component can satisfy multiple capacity bundles and where a stockout clearly loses installer orders.

AY / virtual bundles

Sell 10, 15 and 20 kWh as BOM configurations where possible.

The website may show complete customer solutions while procurement holds common underlying modules rather than one finished SKU for every capacity.

BY / selective options

Keep backup accessories and less common power ratings lean.

Stock only when the verified lead-time penalty and lost-gross-profit risk exceed financing, markdown and obsolescence cost.

CZ / long tail

24–32 kWh residential should usually be order-led.

These configurations can be assembled from core modules or ordered against a committed sale if supplier architecture allows.

Project stock

57–100 kWh light C&I should be customer-funded where possible.

Customer deposits, milestone invoicing and supplier payment terms should materially finance the cash gap rather than Ecodepot warehousing complete C&I systems.

Risk signal

Price compression makes excess stock a margin exposure.

IEA reports average BESS prices in 2025 at roughly one-third of 2020 levels and LFP battery prices down more than 15% during 2025. S09

Do not set “90 days of stock” as a universal rule.

A stock-days target is only defensible after Ecodepot combines actual weekly demand variance, supplier lead-time variance, customer service-level requirement, minimum order quantities, expected price erosion and financing cost. A 90-day rule can be much too high for a rapidly depreciating finished system and much too low for a common module with a long replenishment route during a subsidy surge.

Reorder point = expected weekly sell-through × verified replenishment lead time + safety stock
Statistical safety stock = service-factor × weekly demand standard deviation × √(lead-time weeks)

The statistical formula should be used only when demand and lead-time history are sufficiently stable. Before that, safety stock should be a management-controlled pilot range and explicitly labelled as an assumption.

Inventory return should outrank gross margin percentage.

A SKU with a high percentage margin but low turns, heavy support and rapid price depreciation can destroy more value than a lower-margin module that turns quickly and cross-serves multiple configurations. Product reviews should therefore include annual contribution per average euro of inventory capital, markdowns, RMA burden and stockout loss.

SKU inventory return = annual contribution margin / average inventory capital
Cash conversion exposure = inventory days + receivable days − payable days + explicit supplier prepayment / deposit effects

The working-capital method used in this report treats supplier deposits, final payments, customer deposits, receivable terms, logistics, inventory, financing and guarantees as one cash cycle rather than as separate purchasing metrics.

11 / economics

Supplier price is only one line in BESS contribution.

Public information does not disclose Ecodepot's wholesale cost, rebates, installer discounts, warranty cost, logistics or selling margin. This report therefore does not fabricate gross-margin targets or customer paybacks.

Distributor economics

Model full landed contribution.

Net sales ex VAT
− OEM net purchase price
− dangerous-goods inbound / insurance / handling
− EPR / local compliance / registration cost
− warehousing + inventory financing
− channel incentives / delivery
− commissioning / technical-support cost
− expected RMA / warranty reserve
− markdown / obsolescence reserve
= contribution margin
End-customer economics

Model the bill, not the headline spread.

Value of increased self-consumption
+ tariff / peak-management value where applicable
+ quantified resilience value if customer assigns one
− round-trip losses
− degradation / replacement allowance
− incremental financing / service cost
− foregone export / compensation value
= customer economic benefit
A power-price spread is not a battery margin.

Realised storage value depends on state of charge, usable capacity, efficiency, degradation, power limits, tariff treatment, export compensation, control logic and customer load. Ecodepot should not train sales teams to convert a wholesale or retail price difference directly into battery profit or payback.

Installation complexity belongs in the commercial model.

The procurement comparison should capture installer hours, electrical modifications, number of boxes/components, cabling, commissioning steps, software setup, training, first-time commissioning success, callback rate and remote-resolution rate. A product that is €100 cheaper ex-works but generates an additional site visit can be economically inferior even before customer-experience effects are considered.

The evaluation method therefore converts installation complexity into operating economics: labour, travel, commissioning time, callback probability, remote-resolution capability and installer learning burden must be compared alongside hardware cost.

12 / policy and demand catalysts

Public funding creates upside and timing risk at the same time.

The strategy should be subsidy-ready but not subsidy-dependent. Procurement must distinguish an allocated budget from a published call, validated applications, contracted orders and actual disbursement.

Observed / 2026 policy
RON 400m

Romania's Ministry of Environment stated on 21 May 2026 that the approved AFM budget includes a new programme dedicated to batteries for prosumers and energy storage, financed with 400 million lei, with new sessions intended in 2026. S06

Classification
Allocated

This report treats the 400 million lei as announced/allocated programme funding, not as evidence of awarded projects, paid vouchers or installed batteries. As of the 12 Aug 2026 research cut, no primary source reviewed here supports converting that allocation into realised BESS unit demand.

Procurement implication
Optionality

Secure conditional volume capacity, current documentation and rapid replenishment instead of buying speculative finished-system inventory merely because a programme budget exists.

The upside trigger is operational programme design, not the announcement itself.

Ecodepot should move procurement toward the upside scenario only when the programme's final eligibility, beneficiary type, technical requirements, installer process, application window and reimbursement mechanics are known and the first-order flow of qualified customer demand becomes visible. Before that point, supplier allocation should emphasise option value: reserved production, negotiated lead times and limited deposits rather than irrevocable stock.

13 / five-year demand scenarios

Plan for a 31k–110k annual storage-site envelope, not one deterministic forecast.

These scenarios are internal planning models, not third-party forecasts. They deliberately use conservative declining new-prosumer flows rather than extending the 2025–H1 2026 acceleration. They should be recalibrated when December 2026 ANRE data and final AFM programme mechanics are available.

Downside / internal model
170k

Cumulative additional storage-equipped sites over 2027–2031 in a frozen-June-2026 baseline model.

  • New prosumers fall from 60k to 40k per year.
  • Same-year storage attachment rises from 45% to 55%.
  • Retrofit conversion of opening no-storage stock declines from 4% to 3% annually.
Base / internal model
342k

Cumulative additional storage-equipped sites over 2027–2031 under the most decision-useful central planning case.

  • New prosumers fall from 90k to 70k per year.
  • Same-year storage attachment rises from 60% to 72%.
  • Retrofit conversion of opening no-storage stock moderates from 7% to 5% annually.
Upside / internal model
528k

Cumulative additional storage-equipped sites over 2027–2031 if funding, prices and channel capacity sustain stronger conversion.

  • New prosumers fall from 120k to 100k per year.
  • Same-year storage attachment rises from 70% to 82%.
  • Retrofit conversion of opening no-storage stock declines from 10% to 7% annually as the pool is worked down.
Exhibit 03 / internal planning model / systems per year
Base demand remains around 65k–71k storage-system additions a year without requiring continued exponential growth.
2027–2031 / Romania
Romanian storage-equipped prosumer site addition scenarios 2027 to 2031 Downside annual additions: 36,292; 36,090; 34,622; 32,109; 31,095. Base: 70,261; 71,192; 68,750; 66,990; 64,655. Upside: 107,229; 109,606; 107,241; 104,257; 99,777. 0 30k 60k 90k 120k 2027 2028 2029 2030 2031

Model formula: annual storage additions = new prosumers × assumed same-year storage attachment + opening no-storage prosumers × assumed retrofit conversion. Opening reference: ANRE 30 Jun 2026. Forecast values are Solar Industry Romania internal planning scenarios, not reported market forecasts. H2 2026 additions are intentionally excluded; update after ANRE Dec 2026.

Scenario interpretation

Even the downside model sustains approximately 31,000–36,000 storage-equipped prosumer additions per year. That is sufficient to justify professionalised residential storage operations. It is not sufficient to justify any arbitrary Ecodepot stock quantity because addressable installer share, win rate and system mix remain unknown.

The base case deliberately produces a relatively flat annual additions profile. Growth in same-year storage attachment and continuing retrofit offset a gradual decline in assumed new-prosumer installations. This is strategically useful because Ecodepot does not need exponential market growth for a compact storage portfolio to remain relevant.

Structural-break case

A fourth scenario should remain outside the numeric chart because the evidence does not support a defensible magnitude. A broad, rapidly operational AFM battery programme, materially more time-varying retail prices, aggregator/VPP monetisation or a major change in prosumer compensation could alter both volume and optimal sizing. Conversely, programme delay, favourable export compensation, weaker household financing or severe installer/service bottlenecks could lower adoption materially. The portfolio architecture should survive either direction by using modularity rather than volume bets.

14 / Ecodepot demand model

Convert the national scenario into Ecodepot demand only through observed funnel data.

No public evidence supports an Ecodepot market-share assumption. The commercial forecast should be built from installer/account coverage and measured quote conversion, not from assigning a convenient percentage of Romanian TAM.

Ecodepot annual units = eligible Romanian storage-system additions × addressable installer/channel coverage × qualified-opportunity capture × win rate × fulfilment rate
Category units = Ecodepot annual units × observed use-case mix × observed capacity-band mix
Procurement units = category units + safety-stock requirement − opening sellable inventory − confirmed inbound − interchangeable module substitution
Stage 01
Target installers

Count active installer/EPC/reseller accounts with verified residential or SME customer access.

Stage 02
Qualified opportunities

Separate real customer/project demand from price-checking and generic enquiries.

Stage 03
Configured quotes

Record use case, kW, kWh, phase, retrofit/new-build, brand requested and reason for recommendation.

Stage 04
Wins / losses

Capture loss reason: price, brand, stock, compatibility, financing, installer preference, service or project cancellation.

Stage 05
Repeat / service

Track repeat installer orders, commissioning success, callback, RMA and gross profit after support.

The most important missing dataset is inside Ecodepot, not outside it.

A 24-month export of quotes, orders, SKU/BOM, installer account, capacity, inverter, battery, gross sales, purchase cost, discount, lead time, cancellation, return and warranty data would improve the portfolio decision more than another generic European BESS market forecast.

15 / strategic alternatives

The best current option is WHES anchor plus one conditional specialist.

The recommendation is not “multi-brand.” It is a deliberately constrained form of supplier optionality in which every active supplier has a defined commercial job.

Option Coverage Retrofit readiness Working capital Support complexity Supplier concentration Reversibility Decision
A / WHES only High capacity coverage Unknown until Romanian retrofit pathway is validated Best if portfolio is modular Lowest Highest counterparty concentration Medium Viable if WHES closes retrofit/service/commercial DD gaps.
B / WHES + one specialist Very high High if specialist is selected specifically for retrofit/compatibility Controlled Manageable Improved High Recommended.
C / broad multi-brand warehouse Maximum catalogue breadth High nominally Poor without very high turns Highest Lower logo concentration, not necessarily lower upstream concentration Low once stock is purchased Reject until Ecodepot volume and support economics prove distributor-scale need.
D / project-only residential Available but slow Possible Best Lower Flexible High Too weak for core residential if installers value local availability; retain for long-tail/C&I.

When the opposite recommendation wins

A WHES-only strategy becomes superior if Ecodepot verifies a fully compliant new-build and retrofit architecture, strong Romanian/EU technical service, acceptable RMA economics, reliable stock, competitive landed cost, channel protection and no material brand-specific lost-quote problem. In that condition, a second active supplier would add complexity without enough incremental value.

A broad multi-brand strategy becomes superior only after Ecodepot reaches enough volume to support dedicated technical resources, proves that different brands produce incremental wins rather than cannibalise one another, and can maintain acceptable inventory turns across each ecosystem. This is a scale-stage decision, not a 2027 default.

A project-only residential model becomes superior if subsidy-driven volatility becomes extreme, installers accept long fulfilment times, supplier replenishment becomes very fast and stock availability ceases to influence conversion. Management should test these conditions rather than assume local stock is always valuable.

16 / B2B2C operating model

Installer economics and confidence are part of the product.

In B2B2C, the end customer pays for the system but the installer often determines which brands are quoted, bears commissioning friction and faces the customer when hardware fails. Ecodepot's proposition must therefore solve installer risk.

Installer job 01
Select

Fast sizing workflow, approved BOMs, explicit compatible combinations and use-case templates.

Installer job 02
Install

Predictable hardware, common accessories, training, documentation and lower commissioning time.

Installer job 03
Resolve

Technical hotline, remote diagnosis, spares and a clear RMA route that does not leave the installer financing failure.

Installer job 04
Protect

Deal registration, account protection and disciplined OEM engagement so the installer/Ecodepot relationship is not bypassed.

Commercial differentiation should shift from “we sell batteries” to “we make storage easier to deliver.”

As battery modules become less differentiated and global BESS pricing continues to compress, the distributor's durable margin pool is likely to depend more on configuration, availability, technical enablement, financing, service and warranty execution. This does not require Ecodepot to internalise every activity. Manufacturer-provided level-two support, outsourced specialised logistics and partner installation can remain external where service levels and accountability are contractually clear.

The operating model should preserve customer ownership at Ecodepot/installer level. Supplier access to end-customer telemetry, cloud accounts and commissioning portals must be mapped contractually so technical support does not become an uncontrolled direct-sales route.

17 / five-year roadmap

2027 should simplify; 2028–2031 should scale only what survives real operating data.

H2 2026 is the mobilisation period. The five operating years are designed to move Ecodepot from catalogue availability toward a governed storage platform while keeping large capital commitments conditional on measured repeatability.

H2 26

Build the evidence base

Extract 24 months of Ecodepot quote/order/SKU/margin/service data; complete WHES commercial and compliance DD; define one version-controlled compatibility matrix; identify retrofit technical gap; obtain 2027 battery-passport responsibility evidence.

Decision gate: no incremental strategic supplier until the exact gap is evidenced.

2027

Rationalise and launch retrofit

Operate a deliberately narrow core ladder around approximately 5, 10 and 15 kWh modular configurations plus selected hybrid inverter ratings; make 20–32 kWh mostly configure-to-order; launch an approved retrofit proposition; activate at most one specialist supplier if the WHES route fails the gate.

Commercial objective: improve availability and installer confidence without increasing brand count unnecessarily.

2028

Prove repeatability

Scale B2B2C installer accounts in the highest-opportunity counties; standardise training, commissioning and RMA; launch a project-order 57–100 kWh light-C&I offer; use one full year of capacity-band sell-through to reset stock parameters.

Decision gate: add a premium/installer-pull line only if quantified lost gross profit exceeds its incremental stock/support cost.

2029

Move from hardware to managed energy

Prioritise EMS/API capability, dynamic-tariff readiness, remote diagnostics and fleet-level support. Evaluate whether service contracts, energy optimisation or aggregator partnerships create margin beyond hardware. Review sodium-ion technically but keep it outside core inventory unless economics and support become bankable.

Decision gate: digital/service expansion must demonstrate contribution after support cost.

2030

Reduce structural concentration

Use three years of supplier field performance to renegotiate allocation, price protection, spares and warranty. Map upstream cell/BMS/PCS/cloud concentration rather than brand concentration only. Expand stocked C&I components only if project cadence makes local inventory economically superior.

Decision gate: concentration limits should reflect measured failure/recovery economics and supplier dependence.

2031

Operate a post-subsidy-resilient portfolio

Prune long-tail SKUs, retain modular platforms with strong field data, integrate lifecycle/passport/warranty records, and shift differentiation toward service, compatibility and managed self-consumption. Re-run the supplier slate from zero rather than extending legacy relationships automatically.

End state: compact active catalogue, broad qualified database, measured supplier performance and low inventory obsolescence.

18 / immediate execution

The first 90 days should reduce uncertainty before capital is committed.

Inventory expansion is difficult to reverse after product is imported. Data extraction, supplier DD and controlled retrofit pilots are reversible and therefore should precede scale.

0–30 days

Build the operating fact base.

Export historical quotes/orders, margin, supplier lead time, stock age, lost-quote reason, capacity, inverter, installer and warranty data. Map every current WHES SKU to one strategic role. Request complete legal/compliance/warranty/commercial dossier.

31–60 days

Validate retrofit and supplier alternatives.

Obtain a written WHES Romanian retrofit architecture or record the gap. Run technical and commercial qualification with GoodWe/Solis or another justified specialist. Build a controlled installer pilot using only approved combinations.

61–90 days

Set the first evidence-based stock policy.

Establish ABC/XYZ segmentation, service-level target, reorder logic, maximum aged-stock rules and price-protection escalation. Publish standard 5–7, 9–12 and 14–20 kWh commercial bundles constructed from common parts where technically possible.

3–6 months

Build installer repeatability.

Train priority installer accounts, measure first-time commissioning success and support tickets, record actual quote-loss reasons and isolate which capacity band produces contribution rather than only revenue.

6–12 months

Scale only validated categories.

Expand inventory only for components with repeatable sell-through and demonstrated stockout cost. Launch the 57–100 kWh light-C&I route as project-order with interval-load modelling and disciplined payment milestones.

12-month review

Re-underwrite every supplier.

Compare purchase economics with actual RMA, support hours, installation time, markdown, delivery performance and installer retention. Reallocate volume based on contribution after service, not supplier rebate alone.

19 / red-team

The strategy fails if growth is purchased with stock, support liabilities or channel loss.

The strongest case against the recommendation is not that Romanian storage demand disappears. It is that operational complexity and price compression make a fast-growing category economically unattractive for a distributor that carries the wrong risks.

Anchor supplier cannot close Romanian retrofit or compliance gaps.
Material

Early indicator: missing written compatibility, grid documentation, passport responsibility or repeated installer exceptions.

Response: cap new allocation; activate qualified specialist for the affected use case.

Battery prices fall faster than Ecodepot inventory turns.
High impact

Early indicator: supplier replacement cost declines while aged inventory grows and public market prices reset.

Response: reduce finished-system cover, prioritise common modules, use price protection/rotation where available.

AFM programme timing creates a boom-bust order pattern.
Material

Early indicator: inquiry volume detaches from approved/contracted demand; programme guidance or reimbursement is delayed.

Response: reserve supply conditionally; do not equate applications or announcements with sell-through.

Supplier or manufacturer bypasses the channel.
Strategic

Early indicator: direct customer contact, unprotected project pricing, installer complaints or unexplained direct quotations.

Response: enforce deal registration/non-circumvention; reduce allocation if contractual protections fail.

Firmware or product revisions strand battery/inverter stock.
High impact

Early indicator: compatibility matrix changes, BMS revisions, unsupported mixed-generation expansion or certification updates.

Response: revision-level serial tracking; EOL notice obligations; freeze purchase before unresolved revision transition.

Warranty and callback costs erase headline margin.
High impact

Early indicator: rising tickets per 100 installs, low first-time commissioning success, repeated truck rolls or delayed RMA.

Response: retrain, quarantine affected BOM/revision, raise warranty reserve, shift allocation based on net contribution.

Two suppliers provide little real supply-chain diversification.
Structural

Early indicator: shared cell geography, shared controller dependencies, common cloud infrastructure or same logistics bottleneck.

Response: map upstream dependencies; negotiate resilience at component, warranty and logistics layers.

Residential compensation/tariff economics change.
Structural

Early indicator: legal or tariff change materially alters the value of export, self-consumption or time shifting.

Response: recalculate customer economics and resize commercial messages; do not defend legacy payback claims.

B2B sales are driven by market hype rather than measured load value.
Commercial

Early indicator: proposals without interval data, low conversion, systems sized primarily from PV kWp.

Response: require load evidence and an explicit economic/resilience use case before quotation.

Pre-mortem / 2029.

“The strategy failed because Ecodepot interpreted a fast-growing market as permission to buy product. It carried too many battery and inverter generations, could not manage warranty boundaries, supplier price cuts marked down stock faster than gross profit was earned, and installers migrated to suppliers with simpler commissioning and faster RMA.” The roadmap is designed specifically to prevent this failure mode.

20 / adaptive decision system

Management should change the strategy when observable triggers move.

Scenario planning is useful only if it changes behaviour. The following triggers convert the five-year roadmap into a quarterly operating process.

Demand trigger

ANRE storage-stock trajectory

Re-run the scenario model whenever a new ANRE monthly/annual observation materially changes the storage penetration or no-storage retrofit pool.

Owner: Strategy / quarterly
Policy trigger

AFM programme becomes operational

Move from conditional capacity reservation to incremental stock only after eligibility, timing and qualified customer flow are observable.

Owner: Commercial + Procurement
Portfolio trigger

Brand-specific lost quotes

Activate a third brand only when rolling lost-gross-profit evidence demonstrates incremental demand sufficient to cover additional stock and support.

Owner: Product / rolling 2 quarters
Inventory trigger

Aged stock exceeds policy

Freeze replenishment and reprice/rebalance before adding new model generations. Policy threshold should be set from Ecodepot financing and markdown economics.

Owner: Procurement + CFO
Quality trigger

Support burden deteriorates

Escalate any material rise in tickets, truck rolls, commissioning failure or RMA closure time; compare net contribution after support by supplier.

Owner: Technical / monthly
Supplier trigger

Channel protection breach

Stop new discretionary allocation when deal-registration or non-circumvention obligations fail until commercial protection is restored.

Owner: Procurement / immediate
Technology trigger

New chemistry reaches bankable parity

Move sodium-ion from watchlist to pilot only when landed cost, warranty, certification, field support and lifecycle value compete with LFP for a defined use case.

Owner: Product / annual
C&I trigger

Project cadence justifies local stock

Stock 57–100 kWh systems only when measured project cadence and lead-time opportunity cost exceed carrying, markdown and financing cost.

Owner: B2B + CFO
21 / management dashboard

Eight operating metrics are more useful than a large market-share slide.

The dashboard should expose whether growth is producing contribution, installer repeatability and healthy inventory rather than only revenue.

KPI Why it matters Required cut Management use
Qualified opportunities Leading demand indicator superior to website leads. Installer, new/retrofit, residential/C&I, county, capacity band. Forecast procurement and channel coverage.
Quote-to-order conversion Tests offer and installer relevance. Supplier, capacity, brand requested, stock status, loss reason. Identify real portfolio gaps.
Contribution per order Prevents revenue growth from hiding support/logistics cost. Supplier and use case. Allocate volume.
Inventory turns / aged stock Measures capital productivity and markdown risk. Module, inverter, finished kit, revision. Reorder, freeze or liquidate.
Stockout lost gross profit Quantifies the actual value of local availability. SKU / component / supplier. Set service level rather than arbitrary stock days.
First-time commissioning success Converts technical simplicity into economics. Installer, supplier, firmware/revision. Training and supplier performance.
Support/RMA cost per 100 installs Exposes hidden warranty economics. Supplier / model / issue class. Warranty reserve and supplier reallocation.
Repeat installer rate Measures whether B2B2C execution is becoming repeatable. Cohort and product family. Channel investment and account prioritisation.
22 / research gaps

The remaining uncertainty is identifiable and reducible.

Missing evidence has been kept visible rather than replaced with fabricated market shares, supplier scores, margins or battery-size distributions.

Internal commercial data

Required immediately

Quote and order counts, BOM/SKU, kWh, kW, phase, retrofit/new build, installer, county, customer segment, list/net price, purchase cost, discount, loss reason, stockout, cancellation and repeat purchase.

Internal operations data

Required immediately

Lead time, MOQ, opening/closing stock, stock age, service tickets, installation time, commissioning outcome, RMA reason, closure time, freight/labour reimbursement and defective-battery reverse logistics.

Supplier confidential data

Required before award

Territory, exclusivity, rebates, payment, price protection, account protection, non-circumvention, warranty entity, certification dossier, passport/EPR role, spares, training, roadmap and field-failure references.

Unknown / Romanian capacity mix.

No reliable public dataset identified in this research shows the distribution of Romanian residential storage installations by 5, 10, 15, 20 or 30 kWh category. Any supplier presentation claiming a precise Romanian capacity-market share should be treated as unverified until methodology and denominator are supplied.

23 / board conclusion

Ecodepot should underwrite a storage platform, not a five-year product list.

Battery models will change before 2031. The strategic architecture should therefore define what remains stable: use-case bands, supplier roles, qualification gates, compatibility control, service, inventory logic and trigger-based portfolio renewal.

Approve a 2027 active residential portfolio centred on WHES, with stocked common components supporting approximately 5–7, 9–12 and 14–20 kWh configurations; keep 24–32 kWh and 57–100 kWh systems principally order/project based; create a dedicated retrofit proposition; and activate no more than one additional specialist supplier until Ecodepot's own lost-quote and service data prove that broader brand coverage creates incremental contribution.
Source register and method

Primary evidence first; supplier claims treated as self-reported.

Retrieval and research cut: 12 Aug 2026. Material figures identify period, geography and unit near the claim. Vendor documentation is used for published specifications and product architecture, not as independent evidence of reliability or superiority.

S01 / ANRE — prosumer situation at 30 June 2026

Primary regulator dataset. Romania. Connected prosumers, installed PV capacity and prosumers with storage by legal form; county-level prosumer counts.

Open original source

S02 / ANRE — prosumer situation at 31 December 2025

Primary regulator dataset used for six-month stock comparison.

Open original source

S03 / ANRE — annual prosumer monitoring report 2025

Primary regulator report. Storage growth, compensation contracts, energy delivered and banked-energy statistics; published 2026 using 2025 observations.

Open original source

S04 / ANRE — prosumer regulatory and monitoring portal

Primary regulator reference page for monthly monitoring and prosumer framework.

Open original source

S05 / AFM — consolidated photovoltaic programme guide

Primary programme documentation. Used only for programme architecture and minimum storage threshold; funding rules must be checked against the currently applicable call.

Open original source

S06 / Ministry of Environment — AFM 2026 budget announcement

Primary government communication, 21 May 2026. States a new prosumer battery/storage programme with RON 400 million allocation.

Open original source

S07 / Ministry of Energy — updated final NECP / PNIESC

National policy context. Used as system-level storage context, not as a residential sales forecast.

Open original source

S08 / Eurostat — household electricity prices, second half 2025

Primary EU statistics. Romania recorded the largest year-on-year national-currency household-price increase in the comparison, +58.6%.

Open original source

S09 / IEA — Global battery markets are growing strongly, and so are supply risks

Institutional secondary evidence, Feb 2026. LFP/BESS pricing, chemistry share and manufacturing concentration.

Open original source

S10 / IEA — Global Energy Review 2026: battery storage

Institutional secondary evidence. 2025 global battery-storage deployment and chemistry context.

Open original source

S11 / IEA — Sodium-ion battery momentum grows, but challenges remain

Institutional secondary evidence, 2026. Technology readiness, LFP comparison and sodium-ion manufacturing concentration.

Open original source

S12 / EUR-Lex — Regulation (EU) 2023/1542 on batteries and waste batteries

Primary law. Battery categories, domestic stationary storage classification and Article 77 battery-passport obligations.

Open original source

S13 / European Commission — batteries implementation reference

Primary institutional implementation context for EU battery lifecycle regulation.

Open original source

S14 / EUR-Lex — Regulation (EU) 2025/1561

Primary law amending timing of economic-operator battery due-diligence obligations.

Open original source

S15 / IEC — IEC 62619:2022

Primary standards reference. Safety requirements/tests for secondary lithium cells and batteries in industrial applications, including stationary applications.

Open original source

S16 / IEC — IEC 62485-5:2020

Primary standards reference for safety aspects of stationary lithium-ion battery installations.

Open original source

S17 / UNECE — Manual of Tests and Criteria, Revision 8

Primary dangerous-goods reference. Section 38.3 lithium-battery transport testing context.

Open original source

S18 / Ecodepot — current BESS product portfolio

Company self-reported current public portfolio; used for product-range mapping, not independent quality assessment.

Open original source

S19 / WHES — Battery Box high-voltage LFP family

Manufacturer self-reported specification source. 4.99–29.9 kWh family.

Open original source

S20 / WHES — PC-mini 29.9–50 kW / 57–100 kWh

Manufacturer self-reported specification source for light-C&I architecture.

Open original source

S21 / Huawei — LUNA2000 S1

Manufacturer self-reported capacity architecture, 5–21 kWh current family.

Open original source

S22 / Sungrow — SBR064–256 residential battery

Manufacturer self-reported current HV LFP family, 6.4–25.6 kWh.

Open original source

S23 / BYD — Battery-Box

Manufacturer self-reported HVS/HVM/LVS modular capacity and compatibility architecture.

Open original source

S24 / GoodWe — energy-storage and retrofit solutions

Manufacturer self-reported AC-coupled retrofit architecture.

Open original source

S25 / Solis — AC-coupled residential inverter portfolio

Manufacturer self-reported retrofit architecture for existing grid-tied PV systems.

Open original source

S26 / Deye — AI-W5.1-B residential battery

Manufacturer self-reported 5.12 kWh LFP module specification and compatibility documentation.

Open original source
Method and limitations

Research governance applied in this case study is evidence-first: reported facts, derived calculations, assumptions and forecasts are separated; stock and flow measures are not mixed; supplier evaluation includes technical, regulatory, warranty, service, commercial and working-capital dimensions; scenario ranges are tested against downside conditions; and the final recommendation is explicitly red-teamed.

Portfolio architecture follows a role-based supplier model. Each active supplier must solve a defined commercial or technical problem. The active catalogue should remain compact while a wider qualified-supplier database preserves negotiating leverage and optionality. Forecasting uses downside, base and upside cases rather than a blind CAGR extrapolation.

Visual treatment follows the Solar Industry Romania institutional design system: hard-edged editorial grids, Manrope and DM Mono hierarchy, ink-paper-yellow palette, source/date/unit proximity, deliberate dark data surfaces and mobile recomposition. No external research-system citation artefacts are embedded in the published report.

Key limitations: no public Romanian dataset identified for installed residential battery capacity distribution by kWh; no verified Ecodepot internal sales, margin, stock-turn or warranty dataset supplied; no confidential supplier commercial terms reviewed; no independent field-failure comparison completed; AFM 2026 battery funding is treated as announced/allocated rather than awarded or installed demand. Supplier product specifications remain manufacturer self-reported until Ecodepot validates the exact model/revision dossier.

Decision statement
Narrow the active portfolio, productise retrofit, govern suppliers harder than SKUs, and make inventory a measured service-level decision rather than a growth forecast.

Research status
Research cut 12 Aug 2026. Recalculate the demand scenario when ANRE publishes the December 2026 stock and when the final operational terms of the AFM battery programme are available.