---

First, question the number the quote was built on

Nearly every costly solar-plus-storage mistake starts the same way: a battery quotation is prepared before anyone writes down the actual demand. A household may receive a 10 kWh proposal because the salesperson asked "what is your average electricity bill?", while the real constraint — a 9 kW morning draw from electric showers and a kettle — demands a higher battery power rating and a different inverter. The reverse failure is equally common. Commercial buildings with 2,000 kWh of night-time HVAC load receive residential-scale batteries sized from rooftop area, not from the demand curve.
A load profile is a record of power over time, usually across a 24-hour cycle. It captures three values a monthly bill cannot: the daily energy in kilowatt-hours, the peak power in kilowatts, and the duration of that peak. Those three values define the battery's energy capacity, its power electronics and its usable depth of discharge. A quote that skips the profile is not a quote; it is a projection of hope onto a spreadsheet.
The market context makes this discipline urgent. Battery pack prices have fallen to roughly $115/kWh according to BloombergNEF's 2024 survey, and SolarPower Europe counted 447 GWdc of new PV installed globally in 2023. Storage is no longer a niche accessory. Yet the cost of a poorly matched system is also larger than ever: a correctly sized commercial battery can be $400–$800 per usable kWh installed, so an error of 30% in capacity translates into tens of thousands of dollars of either stranded capital or unserved load.

Build the 24-hour curve before touching a calculator

The first task is data collection, and the standard is simple: record the load for at least two weeks at intervals no longer than 15 minutes, ideally one minute for sites with motor starts or rapid EV charging. Utility interval meters, smart-meter home-area network ports, and temporary current-transformer loggers all supply this data. For a retrofit project, the existing distribution board is the best access point; for new buildings, the designer must synthesize a profile from equipment schedules, occupancy patterns and weather-normalised heating and cooling loads.
That raw data then needs sorting into representative days. Separate weekdays from weekends, school holidays from term time, and summer from winter. A heat-pump home in northern Europe may consume three times more electricity in January than in July; a southern Mediterranean office will show the inverse pattern from air-conditioning. When sizing for self-consumption, use the month where the PV/battery combination is most stressed — usually the low-irradiation, high-load season. When sizing for backup, use the load you genuinely need to keep alive, not the load the building draws at 8 a.m. on a normal Tuesday.
Analysing the profile should yield a one-page summary that the whole tender can reference: total daily energy, maximum 1-second and 15-minute power peaks, the duration of the morning and evening peaks, and the baseload at night. This page becomes the first attachment to every RFQ. Suppliers can then dimension the power block separately from the energy block — a distinction that matters because a battery with 20 kWh of capacity but a 5 kW inverter cannot start a 7 kW borehole pump, however large the pack is.

Sizing mathematics: autonomy, depth of discharge and round-trip losses

Once the profile is approved by the client, the capacity calculation follows a formula that suppliers and auditors recognise:
Required battery capacity = (average daily energy × days of autonomy) ÷ (usable depth of discharge × round-trip efficiency)
Worked through a real case: a rural guesthouse consumes 32 kWh per day and wants two days of autonomy. At a design depth of discharge of 90% for LFP chemistry and a round-trip efficiency of 92%, the formula gives (32 × 2) ÷ (0.90 × 0.92) = 77.3 kWh. That is the nameplate capacity to procure, before manufacturer safety margins are applied.
Three further parameters refine the calculation. First, the cycle-life target: most LFP cells are rated for roughly 4,000–6,000 cycles to 80% of initial capacity, but only if the daily discharge stays within the datasheet's DoD. Second, the C-rate, which simply means the power drawn relative to capacity. A 100 kWh battery discharging at 50 kW operates at 0.5C — a comfortable zone for most LFP systems; the same battery discharging at 150 kW for EV fast charging is at 1.5C and will need a larger pack or a different cell design. Third, temperature: capacity tables in most datasheets assume 25°C, and usable energy falls by roughly 10–15% in unheated outdoor cabinets during winter.
Power electronics sizing follows the profile, not the PV array. The inverter must pass the site's maximum simultaneous export or import, whichever is greater, with headroom for motor inrush. For grid-tied hybrid systems, the relevant figure is usually the peak household load minus the PV contribution at that moment; for off-grid systems, the inverter must carry the full peak load on its own, which is why many off-grid designs use a larger inverter than the PV array would suggest. Delivering the battery's full power through the hybrid inverter also requires DC cabling sized for voltage drop — at 48 V, a 10 kW load pulls more than 200 A, and undersized cable here is a chronic source of efficiency loss and heat.

A bill of materials is a risk register, not a shopping list

The phrase "turnkey battery system" hides a procurement reality: the BOM contains at least ten items that each carry their own warranty, lead time and failure mode.

• Battery modules or racks, with their integrated BMS
• A bidirectional inverter or hybrid inverter with grid-forming or grid-following control
• DC isolators, fuses and surge protection between the battery and inverter
• AC switchgear, metering and the CTs that enable export limiting
• Communication links: Modbus, CAN bus or home-energy management gateway
• Thermal management and enclosure, whether indoor cabinet or outdoor container
• Mounting and seismic bracing for the enclosure
• Cable and containment with the correct rating for DC fault current
• Commissioning documentation, as-built drawings and operator training

For a residential installation, many of these are already embedded in a single AC-coupled battery cabinet such as a 5 kW home battery storage unit. For larger commercial systems, each line demands its own decision. The most overlooked line is often the meter and controller configuration, because grid codes increasingly require export limiting, frequency response or reactive power support; if the inverter cannot accept a remote curtailment signal, the whole design must change.
The comparison below shows how a disciplined buyer treats the three areas where vendors routinely diverge.
| Tender parameter | Common vague wording | What the buyer should request in writing |
|------------------|---------------------|-------------------------------------------|
| Capacity basis | "20 kWh battery" | Nameplate kWh, usable kWh and DoD limit at end of warranty |
| Efficiency | ">90% efficient" | Round-trip AC-to-AC efficiency at 25°C, 0.5C, full cycle |
| Power rating | "5 kW inverter" | Continuous kW at 45°C, surge kW and duration in seconds |
| Cycle life | "10-year warranty" | Throughput in MWh over the term, degradation curve and replacement trigger |
| Standards | "certified" | Specific grid standard (e.g. VDE-AR-N 4105, IEEE 1547) and safety standard (e.g. UL 9540, IEC 62619) |
Keeping this table next to the RFQ prevents the single most expensive ambiguity in the market: suppliers quoting nameplate capacity while buyers believe they are purchasing usable capacity.

Writing the RFQ so that three quotes are actually comparable

A solar battery RFQ fails when it arrives as a two-line email asking for "best price on a 50 kWh system". Suppliers respond with whatever configuration maximises their margin. A rigorous RFQ contains a technical schedule with six mandatory attachments: the approved load profile, the single-line diagram, the site's grid connection class and export limit, the temperature range of the equipment room or outdoor location, the required days of autonomy, and the performance warranty term in years or megawatt-hours.
The cover letter should state the evaluation criteria openly: price per usable kWh, delivered efficiency, warranty conditions, delivery lead time, and the supplier's reference installations in the same application. Pricing the full installed system per usable kWh — rather than per battery rack — is the single best way to compare an integrated cabinet against a modular rack system. An "expensive" rack at $350/kWh nameplate can finish cheaper than a "bargain" cabinet at $450/kWh usable once you strip out the unusable bottom 10% and the extra inverter cost.
Market prices in 2024 give buyers a useful sanity check. Residential installed storage in North America commonly lands between $1,000 and $1,500 per rated kWh, according to installer market surveys, while utility-scale turnkey systems are frequently quoted below $500 per kWh. A quotation far outside that band is not a bargain — it is either a different scope of supply, a different chemistry, or a warranty that will be difficult to enforce. Ask each vendor to list the three largest cost components in their bid. The responses reveal whether they are selling cells, engineering or simply a brand name.
For smaller projects, the RFQ can name a specific product family to anchor the comparison. A hybrid inverter such as the VoltCore 10 kW hybrid inverter gives the design a fixed power-electronics reference, allowing battery vendors to compete only on storage chemistry and integration quality. This separation between the power block and the energy block prevents one weak component from forcing a redesign after contract signature.

Real-world profiles that stress-test the method

Three current application types illustrate why this process pays for itself. The first is the South African household designing for load-shedding. The grid fails for up to four hours at a time, several times daily, but the homeowner only wants lighting, internet, refrigeration and a single circuit of plugs. The captured profile might show a modest 0.8 kW average but a 3.5 kW peak when the fridge starts. A designer who sizes from daily energy alone buys too much capacity and too little power; the profile points instead to a small-capacity, higher-C-rate battery that cycles twice a day without violating the DoD limit.
The second case is a German self-consumption system with a heat pump and an electric vehicle. German households now routinely pair a 10–15 kWp array with 10–15 kWh of storage, but the heat pump adds a 5–7 kW overnight block in winter that a PV-generation-derived sizing method completely misses. A profile-based design shifts the battery's purpose: instead of the common 40–60% self-sufficiency target, the owner can pursue 80% by storing daytime PV surplus for the 2 a.m. heat-pump window, recovering value at €0.30–0.40 per kWh compared with grid purchase.
The third profile comes from commercial EV depots. A warehouse with eight AC charging points at 11 kW draws roughly 88 kW in the late afternoon — a load that coincides with solar production but may exceed the utility transformer's spare capacity. Battery sizing from the depot's charging schedule allows the operator to flatten the peak onto a battery and avoid a six-figure transformer upgrade. In these projects, the load profile is not an input to the battery design; it is the business case. The energy arbitrage alone rarely justifies the capex, but the avoided infrastructure cost usually does.

Where the process breaks, and how to avoid it

Three pitfalls appear repeatedly in poorly executed tenders. The first is treating the load profile as static. Refrigeration plants, EV fleets and industrial shifts all change month to month. A defensible design includes a sensitivity check: what happens to the required capacity if the site's daily energy grows by 20% in year three? The answer determines whether the architecture should use modular racks that can be expanded, or a sealed unit that cannot.
The second pitfall is confusing the BMS with the protection system. The battery's internal management system balances cells and monitors temperature, but it does not substitute for external DC fusing, overcurrent protection or the rapid shutdown initiated by the inverter. Several battery-related fires investigated in recent years traced back to missing external protection between the pack and the inverter; the BOM must list this equipment explicitly rather than assuming it is "included."
The third pitfall is ignoring the inverter's ambient derating. A hybrid inverter rated for 10 kW at 25°C may deliver only 8 kW when mounted in a 40°C plant room next to the battery cabinet. That derating matters most in the very commercial applications — warm rooftops, containerised sites, enclosed electrical rooms — where peak demand is highest. The RFQ should require the supplier to state continuous output at the site's maximum ambient temperature, not at the manufacturer's lab temperature. Our practical experience across battery storage projects and inverter installations shows this single line item resolves more contract disputes than any warranty clause.

Turning the method into a repeatable tender process

Project teams that internalise this workflow can move from raw metering data to an issued RFQ inside three working days. Day one collects and cleans the data. Day two produces the profile summary and runs the sizing calculation with two or three autonomy scenarios so the client sees the cost curve of resilience. Day three assembles the technical schedule and issues it to three or more suppliers, with a request for pricing per usable kWh and a declared degradation curve.
The resulting comparison is meaningful because every bidder answered the same question. The winning proposal is not necessarily the lowest price. It is the system whose datasheet degradation matches the site's actual daily throughput, whose inverter architecture fits the existing AC infrastructure, and whose capacity has been verified against the client's real demand rather than assumed from an electricity bill.
No spreadsheet formula can substitute for a good load measurement, and no supplier guarantee can rescue a system designed around an imaginary number. The load profile is the first deliverable of any serious storage project. A quotation that arrives without one should be treated the same way an engineer treats a structural calculation with no site survey: returned, politely, for more information.