Refrigerator Calculator — Methodology
Every formula and assumption used in the fridge ROI calculator. Designed to be auditable.
In one paragraph
For each fridge we get an annual kWh figure — either from the NEA registry (precise), an age × volume estimate (for old fridges not on the registry), or a custom user-entered value. We multiply by tariff (with annual escalation and GST), accumulate the difference between Option A and Option B, and find the year where cumulative savings cover the upfront cost difference.
Inputs
| Variable | Symbol | Default |
|---|---|---|
| Option A annual kWh (mode-dependent) | kA | ~408 (10-yr 400L estimate) |
| Option B annual kWh (NEA list / custom) | kB | 280 (median 4★) |
| Volume (Estimate mode, A) | V | 400 L |
| Age (Estimate mode, A) | a | 10 years |
| Door-opening factor | φ | 1.0 |
| Salvage / disposal (A) | sA, dA | 0 / 0 |
| Price + delivery − rebate (B) | UB | 1500 + 50 − 0 |
| Electricity tariff (SGD/kWh, before GST) | P | 0.2727 |
| Annual tariff escalation | r | 0.03 |
| GST | g | 0.09 |
| Lifespan horizon | L | 12 years |
Constants (Estimate mode)
| Constant | Symbol | Value | Origin |
|---|---|---|---|
| Base kWh per litre per year | k0 |
0.85 kWh/(L·yr) | Slightly above the 3★ median (~0.81) from the current NEA registry, used as a charitable baseline for "old fridge when new" |
| Age-free window | A0 |
5 years | Empirically, modern fridges hold their factory efficiency for ~5 years before seal/compressor wear matters |
| Per-year degradation past A0 | θ |
4% / yr | Combined effect of door seal slack, compressor wear, refrigerant slow-leak, evaporator dust accumulation |
1Annual kWh per option
Three ways to specify Option A. Option B always uses the NEA list value (or a user override).
A1. NEA registry mode
kA = NEA_kWh(model) × φ
A2. Estimate mode (old fridge, no NEA listing)
age_multiplier = 1 + max(0, a − A0) × θ estimated_kWh = k0 × V × age_multiplier × φ kA = estimated_kWh
Defaults: V=400, a=10, φ=1 → age_multiplier = 1 + 5×0.04 = 1.20 → 0.85 × 400 × 1.20 = 408 kWh/yr.
A3. Custom mode
kA = user_kWh × φ
2Year-N tariff & cost
P(N) = P × (1 + r)N−1 × (1 + g) A_cost(N) = kA × P(N) B_cost(N) = kB × P(N) Savings(N) = A_cost(N) − B_cost(N)
3Net upfront, cumulative savings, break-even
UA = − sA − dA (negative = money in pocket if you keep A) UB = priceB + deliveryB − rebateB Net = UB − UA (extra cash to switch to B) Cum(N) = Σk=1..N Savings(k) Net pos(N) = Cum(N) − Net Break-even year = first N where Cum(N) ≥ Net
We refine the headline year with linear interpolation inside that year so the result reads as a fractional year (e.g. "12.4 yrs"). Special results: Net ≤ 0 → "already paid back"; Savings(1) ≤ 0 → "no payback"; doesn't break even within L → "> L yrs".
Worked example (default inputs)
10-year-old 400L existing fridge (estimate mode), replacing with a typical 4★ 400L unit at 280 kWh/yr. Tariff $0.2727 + 9% GST + 3% annual escalation. Lifespan horizon 12 yrs.
| Step | Calculation | Result |
|---|---|---|
| kA (Estimate) | 0.85 × 400 × (1 + 5×0.04) × 1.0 | 408 kWh/yr |
| kB | NEA listed 4★ value | 280 kWh/yr |
| Year-1 tariff | 0.2727 × 1.09 | $0.2972 / kWh |
| Year-1 A cost | 408 × $0.2972 | $121 |
| Year-1 B cost | 280 × $0.2972 | $83 |
| Year-1 savings | 121 − 83 | $38 |
| UA | −0 − 0 | $0 |
| UB | 1500 + 50 − 0 | $1,550 |
| Net upfront | 1550 − 0 | $1,550 |
| 12-yr cumulative savings | ~$38 × ((1.03)12−1)/0.03 | ~$540 |
| Payback | — | > 12 yrs (doesn't pay back on energy alone) |
Assumptions & limitations
- NEA Annual kWh is reasonable. Unlike aircon (where NEA's standard usage hours barely match real life), fridges run 24/7 at standardised 25°C ambient — close to SG kitchen conditions. Real-world variance vs label is usually under 15%.
- Door-opening factor is a single multiplier. A 4-person household opens a fridge 2× more than a 2-person one but that's a tiny effect on annual kWh (~5–10%). Default 1.0 is fine for typical use; bump to 1.1 if very heavy.
- Estimate mode is rough. Old fridges vary wildly — a well-maintained 15-yr-old unit might be at 1.0×, a poorly-maintained 12-yr-old at 1.5×. The age curve gives a sensible middle estimate but isn't a substitute for measuring (smart plug for a week + extrapolate).
- No frost-free / inverter / climate-class adjustments. All built into the NEA published Annual kWh; we don't double-count.
- Tariffs compound annually at constant
r. Real tariffs revise quarterly with double-digit swings. - Servicing not modelled. Fridges rarely need service in their first 8–10 yrs, and after that you're usually replacing anyway.
- Linear interpolation for fractional break-even year. Good to ~1 month accuracy.