Excel vs a feasibility tool: what a spreadsheet misses

What a spreadsheet does well

• It is already paid for, and everyone on the client's side can open it.
• Every formula is visible. A reviewer can follow any number back to its inputs.
• NPV, IRR and payback are built-in functions, and a sensitivity table is a few minutes' work.
• It does exactly what you build into it, including the unusual tariff or contract no tool has.

Zorowatt makes a feasibility tool, so read the rest with that in mind. The point of this article is where spreadsheet studies tend to go wrong, so you can check yours whichever way you work.

Ten things a spreadsheet battery study often misses

1. The size of the year. 365 days × 96 quarter-hours = 35,040 rows; a leap year has 35,136. A formula range that stops short, or a file that starts on 2 January, quietly drops the peak that mattered.
2. Daylight-saving days. In Europe the spring day has 23 hours (92 quarter-hours) and the autumn day 25 (100). Exports handle the doubled autumn hour in different ways: repeated timestamps, a skipped hour, or local winter time all year. Read it wrong and two quarter-hours are added together into a peak that never happened.
3. kW or kWh. A 15-minute value of 100 kWh is an average of 400 kW. A column read as kW when it is kWh per interval puts every peak off by a factor of four.
4. Cumulative registers. Some exports give the meter reading, not the consumption; the consumption is the difference between consecutive readings.
5. Gaps, stuck zeros and flat lines. A day of zeros looks like a cheap day; a stuck value looks like baseload. Both change the size.
6. Which peak the tariff bills. In the Netherlands the kWmax is the highest quarter-hour of each month; in Germany the Leistungspreis is set by the year's single highest quarter-hour, so one missed day costs the year.
7. The 2,500-hour switch in Germany. Full-load hours are annual kWh ÷ annual peak kW. A battery that cuts the peak raises them, and a site just below 2,500 hours can cross into the band with the higher kW price and the lower kWh price. A spreadsheet that applies one price pair misses it.
8. Recharging. The battery can only refill in the room below the target peak (or below the connection limit) before the next peak. Without a quarter-hour state of charge, two peaks three hours apart look as easy as one.
9. Losses on the bill. Every kWh lost in the battery is bought, with its network charge and energy tax. At 90% round trip, delivering 22,500 kWh a year means buying 25,000.
10. Checking the bill. If last year's data, run through the spreadsheet's tariff, does not reproduce last year's invoices, the savings figure has nothing to stand on. Typing in twelve invoices line by line is the step most often skipped.

More on each data problem: the ten data problems that size a battery for you. On the German tariff: German grid charges and the 2,500-hour rule.

If you stay with the spreadsheet

A checklist that catches most of the above:

• Count the rows per day: 96, with 92 and 100 on the daylight-saving days (if the export follows local time).
• Compare the year's kWh with the invoices. A factor of about four means kW and kWh are swapped.
• Find the twenty highest quarter-hours and look at the days around them for duplicates or spikes.
• Count the share of real readings; Zorowatt treats 12 months with at least 90% measured as the minimum for a sound study.
• Rebuild one month's bill from the data and compare every line with the invoice before you model the battery.
• Simulate the battery per quarter-hour with a state of charge that cannot go below zero or above the usable capacity.
• Bill the losses, and model degradation and the replacement year in the cash flow.

Side by side

The full comparison, with the rows for every tool: Excel vs Zorowatt.

What the time is worth

The research behind this article found no published fixed fees for C&I battery studies in Germany, the Netherlands, the UK or Romania. Zorowatt's own estimate, not a measured figure: a simple study (one site, 12 months of 15-minute data, peak shaving plus PV self-consumption, NPV) takes 2–5 engineer-days. If a tool takes a day or two off that, compare it with your day rate and the number of studies you do a month. If you do two studies a year, the spreadsheet is probably the cheaper choice.

Ready-made spreadsheet models

Battery financial models are sold on template marketplaces. Their prices could not be read directly when Zorowatt's research checked them (the pages returned an error), so they are not quoted here. Before buying one, check whether it reads 15-minute meter data and simulates a battery per interval, or whether it starts from annual figures.

Getting numbers out to Excel

A tool does not have to replace the spreadsheet. The Workspace exports results as CSV, so the cash flow can go into the client's own model, and the PDF report lists the tariff versions and assumptions so every number can be retraced. For a first peak-shaving number without either, use the free peak-shaving calculator.

Sources

• NEW Netz, Preisblatt Strom 2026 (Leistungspreis on the annual peak, 2,500-hour split)
• Stedin, Vergoedingsregeling elektriciteit 2024 (kWmax per clock quarter-hour)
• Liander, tariffs for connection and transport, large consumers 2026

Last checked: 2026-10-30. Facts about other tools were read on 2026-10-28 from the pages linked next to them; prices and features change, so check the vendor's page before you rely on one. Examples marked illustrative are not real customers. The study-effort figure is an estimate; the arithmetic is shown where it is used.