A data-driven look at my residential solar setup — what works, what doesn’t, and why the cheapest upgrade might be the smartest.
The Setup
In February 2024, I installed a 5kW on-grid solar system at my home in Bangkok — 9 panels with a Huawei SUN2000-5KTL-M2 inverter. Total cost: about 150,000 THB. The system feeds into a 3-phase electrical system and sells excess power back to MEA at 2.20 THB/unit.
After a year of living with it, I wanted to understand: how much is it actually saving me? And should I upgrade?
The Core Problem
Here’s the uncomfortable truth I discovered after mapping my household’s consumption pattern:
| Time of Day | Solar Output | Who’s Home | What’s Running |
|---|---|---|---|
| 07:30–16:00 | Peak solar | Nobody | Fridge, NAS, routers (minimal) |
| 16:00–19:00 | Fading | Family returns | 1st floor AC (29K BTU) |
| 19:00–06:00 | Zero | Everyone | 2nd floor ACs, EV charging, water heaters |
Peak solar hours = nobody home. Peak consumption = no solar.
About 85–90% of my electricity consumption happens when the sun isn’t shining. The solar panels spend most of the day generating power that gets exported at 2.20 THB/unit, while I buy it back in the evening at 3.25–4.42 THB/unit.
The 3-Phase Balance Problem
There’s a second hidden issue. My inverter (SUN2000-5KTL-M2) is a balanced-phase model — it distributes solar output equally across all 3 phases.
When only one phase has a heavy load (like an AC unit), the solar on the other two phases gets exported at 2.20 THB while the loaded phase simultaneously buys from the grid at ~4.00 THB. You’re selling cheap and buying expensive at the same time.
What the Bills Actually Say
I tracked 6 months of MEA bills:
| Period | Grid Purchase (kWh) | Solar Sold (kWh) | Net Payment (THB) |
|---|---|---|---|
| Jul–Aug 2025 | 1,203 | 348 | — |
| Aug–Sep 2025 | 1,240 | 417 | — |
| Sep–Oct 2025 | 1,111 | 367 | 4,428 |
| Oct–Nov 2025 | 1,071 | 305 | 4,368 |
| Nov–Dec 2025 | 916 | 373 | 3,460 |
| Dec–Jan 2026 | 922 | 351 | 2,697 |
Average net payment: ~3,738 THB/month with solar.
Before solar, I was paying somewhere in the 5,000–7,000 THB range (from memory — I don’t have the old bills). That puts monthly savings at roughly 1,200–2,300 THB, with payback somewhere in the 5.5–9.9 year range.
Honest answer? I can’t be more precise without pre-solar bills.
The Upgrade Options
I evaluated five paths forward:
Option A: Switch to TOU Tariff — 7,340 THB
This is the cheapest and smartest first move. MEA’s Time-of-Use (Rate 1.3) charges:
- Peak: 5.80 THB/unit (Mon–Fri 09:00–22:00)
- Off-peak: 2.64 THB/unit (Mon–Fri 22:00–09:00 + all weekends + holidays)
Why this works for my house:
- Solar covers peak daytime hours (09:00–16:00) → minimal expensive grid purchase
- Sleeping ACs run off-peak (22:00–06:00) → cheap rate
- All weekend consumption is off-peak — every single unit at 2.64
- EV charging overnight → off-peak
My estimated peak/off-peak split: 20–25% peak / 75–80% off-peak.
| Scenario | Monthly Savings | Payback |
|---|---|---|
| Optimistic (20/80) | 851 THB | ~9 months |
| Base case (25/75) | 681 THB | ~11 months |
| Conservative (30/70) | 511 THB | ~14 months |
Risk: Low. If it doesn’t save as expected, you can switch back to progressive rate.
Option B: Upgrade to 10kW — 124,000 THB
Replace the inverter with Huawei SUN2000-10K-MAP0. The key advantage isn’t just more power — it’s per-phase optimization. The new inverter routes solar to whichever phase has load, eliminating the simultaneous buy/sell waste.
Also: it’s battery-ready for the future (supports LUNA S1 directly).
Option C: Battery Only — 184,710 THB
Huawei LUNA S1, 14 kWh. Not recommended alone — with only 5kW generation and the 3-phase balance issue, there isn’t enough surplus solar to fill a 14kWh battery.
Option D: TOU + 10kW — 131,340 THB
The power combination. TOU makes off-peak cheap, 10kW solar covers peak, phase optimization eliminates waste. Estimated payback: ~3–3.5 years.
Option E: Everything — 316,050 THB
TOU + 10kW + Battery. Marginal benefit from battery is low because TOU already makes off-peak cheap at 2.64 — the arbitrage from battery storage is less compelling.
The Comparison
| Option | Investment | Monthly Savings | Payback | Risk |
|---|---|---|---|---|
| A: TOU only | 7,340 | 680–850 | 9–11 months | Low |
| B: 10kW only | 124,000 | ~2,200 | ~5 years | Medium |
| C: Battery only | 184,710 | ~300–600 | 25–35 years | High |
| D: TOU + 10kW | 131,340 | ~3,000–3,700 | ~3–3.5 years | Medium |
| E: All three | 316,050 | ~3,300–4,000 | ~7–8 years | High |
My Plan
Step 1 (NOW): Switch to TOU — 7,340 THB
Quick win. Validates actual peak/off-peak data.
Step 2 (AFTER): Upgrade to 10kW — 124,000 THB
Combined payback ~3–3.5 years.
Step 3 (LATER): Battery — only if prices drop or conditions change.
Why this sequence:
- TOU first = cheapest, fastest payback, reversible, provides real data
- 10kW second = validated by TOU data, fixes phase balance, doubles generation
- Battery last = lowest marginal value when TOU already makes off-peak cheap
What I Learned
- Track your actual data. Without bills, you’re guessing. I wish I had kept my pre-solar bills.
- The mismatch matters more than capacity. Doubling solar from 5kW to 10kW doesn’t double savings if nobody’s home during peak generation.
- Phase balance is invisible but expensive. The balanced-phase inverter was silently wasting money by selling cheap and buying expensive simultaneously.
- TOU is the hidden gem. A 7,340 THB meter swap can save more per baht invested than a 184,710 THB battery.
- Sequence your investments. Each step generates data that validates (or invalidates) the next step.
This is a living note. I’ll update it as I switch to TOU and collect real data.
Status: Step 2 analysis complete. TOU switch pending.