Sizing a small energy system is not one calculation but a short sequence, each step feeding the next: understand the use case, measure the loads, check power delivery, size storage, size generation, then add visible margins and iterate. This guide runs the whole sequence and threads one worked example through it, so you can see how the numbers connect. It is the pillar for the system-design cluster; each step links to the article that covers it in depth.

Two boundaries up front. First, everything here produces a preliminary, educational estimate — a starting design to reason about and refine, not an engineered, code-compliant installation. Conductor and protection sizing, grounding, and any grid connection are governed by your local standard and often require a qualified professional. Second, the method is deliberately built from reusable formulas and your own local inputs — solar resource, battery chemistry, prices — rather than fixed regional numbers, so it stays valid wherever and whenever you use it.

The sizing sequence at a glance

1. Use case & architecture  →  what are we powering, and in what kind of system?
2. Daily energy demand      →  how many Wh per day?              (load profile)
3. Power delivery           →  continuous / peak / surge W?      (power article)
4. Storage                  →  usable & nominal kWh, then Ah at V
5. Generation & charging    →  replace the daily Wh under real conditions
6. Margins & iterate        →  visible reserves; loop back if a step forces a change

Steps 2 and 3 are independent questions about the same loads — energy and power — and both must be answered. Steps 4 and 5 depend on them.

Step 1: use case and architecture

Start where understanding small energy systems leaves off: state what the system powers, and which architecture it is — grid-connected, backup, mobile, or standalone. This decides which parts you need, how much autonomy matters, and which local rules apply. A standalone cabin lives or dies by its autonomy through poor weather; a grid-tied backup mainly needs to carry a chosen set of loads for a defined outage.

Step 2: daily energy demand

Build a daily energy load profile. That article's method — measuring cycling loads over time, counting standby and inverter idle draws, and adjusting AC loads for inverter losses — produces the single number the rest of this depends on.

Worked example. We will carry forward that article's result:

Daily energy demand  E_daily ≈ 1.0 kWh/day  (illustrative)

This figure already includes the inverter's idle draw and any AC-load conversion losses, so it represents energy drawn from the battery.

Step 3: power delivery

Before sizing storage, settle the power question from continuous power, peak power, and surge loads, because it sizes the inverter and the delivery path independently of energy:

  • Continuous: the sum of loads that run for long stretches.
  • Peak: the largest realistic sum of loads running simultaneously.
  • Surge: the biggest single startup surge (usually the largest motor) — check it against the inverter's surge rating, not its continuous rating.

Carry forward that article's example: a pump running at 600 W with a startup surge near 1 800 W needs an inverter comfortably above 1 000 W continuous and 2 000 W surge, and a delivery path (battery/BMS, cables, protection) that can pass the surge current. Remember that at 12 V an 1 800 W surge is about 150 A, so the battery's discharge limit and cabling matter as much as the inverter.

Step 4: size the storage

Storage sizing converts daily energy and an autonomy target into a battery size. The core relationship:

Usable energy needed  = E_daily × N            (N = days of autonomy)
Nominal energy needed = Usable energy ÷ u      (u = usable-capacity fraction)

Days of autonomy (N) is how long the system must run without meaningful charging — a design choice driven by your use case and how tolerant it is of a shortfall. Usable fraction (u) captures how much of a battery's rated energy you can actually use, after depth-of-discharge limits and reserve. It depends heavily on chemistry, and temperature and aging reduce it further — so treat u as an input you take from usable vs. nominal battery capacity and your battery's data sheet, not a universal constant.

Continuing the example, with an illustrative N = 2 days and u = 0.8:

Usable energy  = 1.0 kWh × 2      = 2.0 kWh
Nominal energy = 2.0 kWh ÷ 0.8    = 2.5 kWh  (nominal battery rating)

The usable fraction matters enormously: at u = 0.5 the same demand needs a 4.0 kWh nominal battery instead of 2.5 kWh. Getting u right — from real chemistry limits, not optimism — is often the difference between a system that lasts and one that ages prematurely.

To translate nominal energy into a battery rating in amp-hours, divide by the system voltage (a nominal figure — see watt-hours vs. amp-hours):

Ah = Wh ÷ V   →   2 500 Wh ÷ 12 V ≈ 208 Ah
                  2 500 Wh ÷ 24 V ≈ 104 Ah

The same energy needs roughly half the amp-hours at 24 V that it does at 12 V — and, from Step 3, draws half the current for a given power. This is why system voltage is a genuine design decision, not a detail.

Step 5: size the generation and charging

Generation has to replace the daily energy the loads remove — under realistic conditions, not average ones. For solar, the daily energy a panel produces is roughly:

Solar Wh/day ≈ Panel watts × peak-sun-hours × system derate

where peak-sun-hours (PSH) is a location- and season-specific measure of usable sunshine, and the derate (well below 1) bundles temperature, soiling, controller, and wiring losses. Rearranging to size the array:

Panel watts ≈ E_daily ÷ (PSH × derate)

With illustrative PSH = 4 h and derate = 0.7:

Panel watts ≈ 1 000 Wh ÷ (4 × 0.7) = 1 000 ÷ 2.8 ≈ 357 W  → round up to ~400 W

The critical caveat is which PSH you use. Sizing to the annual average leaves a standalone system short every winter; the same demand at PSH = 2 h needs about 714 W, nearly double. Design generation for the worst month you must operate through, and confirm your local solar resource and derate before trusting any figure — the estimation itself belongs in the power sources cluster, with your location's data. Other sources (alternator, shore power, a generator) are sized the same way: can they replace the daily energy in the charging time actually available?

Step 6: margins, and iterate

Real systems need visible reserves — and each should be a decision you can point to, not a fudge hidden in the loads:

  • Energy margin: the days-of-autonomy N and any allowance for higher-use days and battery aging.
  • Power margin: headroom above continuous and surge demand so the inverter is not running at its limit.
  • Generation margin: sizing to a poor month, plus derate, rather than to ideal conditions.

Sizing is iterative. If Step 5 shows the array is impractically large, you revisit Step 2 (can a load be reduced, or run on DC?) or Step 1 (does the use case need a supplementary charging source?). A good design is usually the second or third pass, not the first.

Worked design, gathered

For the illustrative ~1.0 kWh/day case, with the assumptions stated above:

Quantity Result Key assumptions (yours will differ)
Daily energy ≈ 1.0 kWh/day includes inverter idle & AC losses
Inverter ≥ 1 000 W cont. / 2 000 W surge sized by the pump's surge, not energy
Storage (usable) 2.0 kWh N = 2 days autonomy
Storage (nominal) 2.5 kWh (≈ 208 Ah at 12 V) u = 0.8 usable fraction
Solar array ≈ 400 W PSH = 4 h, derate 0.7 (average month)

Every figure here is a preliminary estimate tied to visible assumptions. Change an assumption — a lower usable fraction, a winter PSH, an extra load — and the design moves in a direction you can trace. That traceability is the point: a number you can defend and revise beats a number handed down by a black box. (When you want to run many such scenarios and keep them, that is exactly the repeated workflow a Viherion planner is meant to own; the method itself stays public and usable by hand.)

Before you build

  • Confirm battery chemistry limits, and conductor and protection sizing, against the manufacturer and your governing standard — see why protection is part of system design and the safety fundamentals.
  • Treat grid connection, battery assembly, and any energized work as tasks for a qualified professional where your jurisdiction requires it.
  • Stop and reassess at any sign of heat, arcing, swelling, odor, or unexplained protection operation.

This sizing math tells you roughly how big the pieces should be. It does not certify that an installation is safe or compliant — that judgment belongs to the current standards and, where needed, a qualified professional.

Key takeaways

  • Sizing is a sequence: use case → daily energy → power delivery → storage → generation → margins, iterated.
  • Energy sizes the storage and generation; power (continuous, peak, surge) sizes the inverter and delivery path. They are separate.
  • Nominal storage = (daily energy × days of autonomy) ÷ usable fraction; the usable fraction, from real chemistry limits, drives the result.
  • Size generation for a realistic worst month, not the annual average.
  • Keep every margin visible; expect to iterate.
  • The output is a preliminary estimate — confirm chemistry, protection, and compliance locally and with qualified help.

Where to go next