Writing

REPS: simulating which retrofit policy actually reaches Indonesia's Type 36 homes

A browser-based policy simulator that scores subsidies, green finance, mandates, and behavioural nudges on adoption, cost-effectiveness, and equity — built on the engineering results from my net-zero Type 36 thesis.

My net-zero Type 36 thesis showed that a low-cost "Cooling Survival Kit" can cut simulated electricity use by ~57%. The next question — the one policymakers actually ask — is which instrument gets households to adopt it?

REPS (Residential Retrofit Policy Simulator) is a transparent, browser-based prototype I built to explore that question. It couples the engineering savings from my Energy3D work to a discrete-choice (logit) model of household adoption, then lets you pull policy levers and watch the trade-offs unfold.

What you can do with it

Open the interactive simulator → and try:

  • Consumer subsidies — how much of the retrofit cost the government covers
  • Green financing — subsidised loans that remove upfront liquidity constraints
  • Construction mandates — requiring the kit on new builds
  • Behavioural nudges — campaigns that offset adoption friction

Sliders also expose the assumptions that matter: present-biased discount rates, rebound effects (savings taken as comfort rather than cash), rationed-comfort welfare for poorer deciles, and adoption friction — the behavioural barriers behind the energy-efficiency gap.

What it outputs

For any policy mix, REPS aggregates over 10 income deciles (illustrative Indonesian urban ranges, or your own CSV import):

  • Adoption rate and households retrofitted
  • Electricity and CO₂ savings
  • Fiscal cost, cost per ton CO₂, and cost per household
  • Equity charts — who adopts and who benefits (the poor gain most in comfort, the rich in cash)
  • A four-instrument comparison table — mandate vs subsidy vs green finance vs nudge — with no single winner on every criterion

The model, honestly

Per household, yearly net benefit is:

NB = cash saving + value of recovered comfort − annualised cost of the retrofit

Adoption follows a logit on a benefit-vs-affordability index, penalised by friction and liquidity constraints, boosted by nudges, and forced on new builds under a mandate. Parameters come from my thesis (57% saving, ~IDR 8M retrofit cost), PLN tariffs, and IEA/ESDM grid emission factors.

Caveats I'll state up front: the logit scale is assumed, not econometrically estimated; income deciles are representative, not survey data; comfort valuation is a parameter, not measured. Everything is adjustable so the sensitivity is visible. This is a decision-support prototype — the doctoral step would be estimating preferences from revealed-choice data.

Why it matters

Indonesia's "One Million Houses" program locks millions of households into thermally poor envelopes. My thesis showed the engineering case for a cheap retrofit; REPS shows the policy design case — which lever reaches the most households, at what fiscal cost, and whether the poorest deciles are left behind.

Try the simulator → · Read the net-zero Type 36 summary → · Full paper on ResearchGate →