Energy: PEA tariff, electric vs biomass steam, PV vs solar thermal

PEA tariff schedule (May 2023 onward, still current) filed under
docs/finance/pea/; Ft 0.1623 THB/kWh for Sep-Dec 2026 from the PEA Ft page.
The plant falls under Schedule 3.2.2 (Medium General Service, TOU, 22 kV).

New section in sustainability.md: blended ~3.81 THB/kWh for two-shift
production plus 132.93 THB/kW demand charge; electric steam ~4.7M-5.5M
THB/yr at maturity against ~0.35M-0.8M of biomass fuel; self-consumed PV
worth ~3.7 THB/kWh (~4.5-7 yr payback); solar thermal not recommended;
the battery earns little from TOU arbitrage or peak shaving. biomass.md,
scale.md, the utilities CSV note and the README follow.

docs/sustainability.md

@@ -27,14 +27,19 @@ at 10 bar / 180 °C) and the 250 kW cooling system.
   rack's DC voltage window to the specific Deye model (a full 15-pack rack
   can exceed the 800 V window of the smaller Deye series), and obtaining
   Gobel's safety certifications (not published) for a Thai commercial
-  install. At 271 kWh the stack is a peak-shaving and ride-through buffer,
-  not full off-grid autonomy, against loads like the 250 kW cooling system.
+  install. At 271 kWh the stack is a ride-through buffer, not full
+  off-grid autonomy, against loads like the 250 kW cooling system; under
+  PEA's time-of-use tariff it earns little from peak shaving
+  ([grid tariff](#grid-tariff-pea-and-what-it-decides)).
 - The machine load and cooling system are straightforward PV targets.
-- Steam is the open engineering question. Options under evaluation include an
-  electric boiler (maximizes the solar share), a biomass boiler (locally
-  abundant fuel), and solar-thermal pre-heating of boiler feedwater. No
-  technology choice has been made yet; it will be driven by honest numbers on
-  load profiles, storage, and grid backup, not by what sounds greenest.
+- Steam is the open engineering question. Options under evaluation are an
+  electric boiler (maximizes the solar share) and a biomass boiler (locally
+  abundant fuel); solar-thermal pre-heating of boiler feedwater was
+  considered and is not recommended. No technology choice has been made
+  yet; it will be driven by honest numbers on load profiles, storage, and
+  grid backup, not by what sounds greenest. At PEA's tariff, biomass heat
+  costs about a tenth of electric steam
+  ([grid tariff](#grid-tariff-pea-and-what-it-decides)).
 - Whole-factory HVAC (continuous temperature, humidity, and PM control, see
   [facility systems](facility-security.md)) adds a major tropical-climate
   load on top of process cooling and must be part of the PV and battery
@@ -117,9 +122,10 @@ reusing it before generating anything new:
   dehumidification reheat (the 35 to 60% RH spec means air is cooled below
   dew point and reheated continuously) run on recovered press heat instead
   of fresh electricity.
-- **Solar thermal feedwater preheat**: flat-plate or evacuated-tube
-  collectors reach 60 to 90 °C trivially in the Thai sun; cheap, proven, and
-  stackable with the heat pump recovery.
+- **Solar thermal feedwater preheat** (considered, not recommended):
+  flat-plate or evacuated-tube collectors reach 60 to 90 °C easily in the
+  Thai sun, but that is the temperature the heat pump already recovers for
+  free, and the roof earns more as PV (see the grid tariff section below).
 - **Steam accumulator**: the presses draw steam in bursts, so an accumulator
   (already a Newbilt option) lets a smaller boiler ride the peaks.
 - **The boiler only does the final lift**: whichever technology is chosen
@@ -145,6 +151,94 @@ diesel means a fuel tank with price exposure. The electric and biomass paths
 avoid a fossil tank on the property entirely, which matters for both the
 sustainability commitments and the BOI story.
 
+### Grid tariff (PEA) and what it decides
+
+Source: PEA's electricity tariff schedule effective from the May 2023 bill
+(unofficial English translation, filed as
+[`finance/pea/EN_Electricity_Tariffs_May_2023.pdf`](finance/pea/EN_Electricity_Tariffs_May_2023.pdf)),
+still the current schedule on PEA's
+[tariff page](https://www.pea.co.th/en/our-services/tariff), and the Ft on
+PEA's [Ft page](https://www.pea.co.th/en/our-services/tariff/ft), both
+checked 7 October 2026. Plant consumption and demand are this repo's
+estimates ([scale](scale.md)).
+
+**Which tariff applies.** At 300 to 750 kW of peak demand the plant is
+Schedule 3, Medium General Service (30 to under 1,000 kW and at most
+250,000 kWh a month on a three-month average). The flat-rate 3.1 is closed
+to customers classified after November 2015, so a new plant is on **3.2,
+Time of Use**, and with a 22 kV connection on 3.2.2:
+
+| Item | Rate |
+|---|---|
+| Demand charge | 132.93 THB per kW of the highest 15-minute demand in peak hours |
+| Energy, peak (09:00 to 22:00, Monday to Friday) | 4.1839 THB/kWh |
+| Energy, off-peak (22:00 to 09:00 on weekdays; all day on weekends and public holidays) | 2.6037 THB/kWh |
+| Service charge | 312.24 THB/month |
+| Ft, fuel adjustment (September to December 2026) | +0.1623 THB/kWh, reset every four months |
+| VAT | 7 % on top; reclaimable for a VAT-registered company, so left out below |
+
+Rules that shape the design: metering on the low-voltage side of our own
+transformer adds 2 % to kW and kWh, so meter at 22 kV; the minimum bill is
+70 % of the highest demand charge of the last twelve months; a power factor
+below about 0.85 (reactive demand above 61.97 % of active) costs 56.07 THB
+per kVAR, so the motor loads need correction. With electric steam the plant
+would average ~210,000 kWh a month (estimate) and busy months could cross
+250,000, which moves it to Schedule 4; that schedule's TOU rate 4.2.2 is
+identical, so the price does not change. Connecting a high-voltage meter
+below 69 kV costs a 15,000 THB inspection fee plus a security deposit of
+400 THB per kVA (PEA tariff page, rates from January 2020): 200k to 400k
+THB for a 500 to 1,000 kVA transformer.
+
+**What it costs the plant.** Two shifts from 06:00 to 22:00 on ~300 days
+(about 243 weekdays and 57 Saturdays, estimate) put ~66 % of process energy
+into peak hours: a blended **~3.81 THB/kWh** including Ft, before the
+demand charge. Without electric steam (1,000 to 1,500 MWh a year, 300 to
+450 kW peak) the bill comes to roughly **4.3M to 6.4M THB a year** before
+solar, 0.5M to 0.7M of it demand charge (estimate).
+
+**Electric versus biomass steam**, for the same heat as in
+[biomass](biomass.md): 350 to 400 t of wood at 80 % boiler efficiency,
+~1,120 to 1,280 MWh of heat a year at maturity.
+
+| | Electric boiler (~98 %) | Biomass boiler (~80 %) |
+|---|---|---|
+| Price of heat | ~3.9 THB/kWh, plus the demand charge | ~0.3 to 0.6 THB/kWh at 1 to 2 THB/kg of fuel (fuel price not yet checked with local suppliers) |
+| Running cost, mature two shifts | **~4.7M to 5.5M THB/yr**: 1,140 to 1,310 MWh plus 200 to 300 kW of extra peak demand (0.3M to 0.5M) | **~0.35M to 0.8M THB/yr** of fuel, plus fuel handling and ash |
+| Boiler package ([utilities CSV](finance/utilities-investment.csv)) | 2.5M to 4M THB | 4M to 7M THB, plus the fuel store |
+| Grid | peak demand 600 to 750 kW: bigger transformer and deposit | unchanged |
+
+The electric boiler is the more efficient machine; biomass is by far the
+cheaper heat. At maturity the fuel saving of ~4M to 5M THB a year repays
+the biomass premium in about a year (estimate). Even early, at a quarter of
+the heat, electric steam costs ~1.4M to 1.7M THB a year against ~0.1M to
+0.2M of fuel, because the demand charge is set by the boiler's full-power
+draw, not by its average. Moving electric steam to off-peak hours barely
+helps: the presses need steam while they run, and the peak window covers
+13 of the 16 production hours. The staged electric-first path in
+[biomass](biomass.md) is therefore a permitting and fuel-logistics choice,
+not a cost one.
+
+**PV, not solar thermal.** Self-consumed PV is worth **~3.7 THB/kWh**
+(weekday daytime at the peak rate, weekends and early mornings at
+off-peak, Ft included), so the 247 kWp roof saves ~1.2M to 1.3M THB a year
+and pays back in **~4.5 to 7 years** at 25,000 to 35,000 THB/kWp
+(estimate). It does not lower the demand charge, because the peak window
+runs to 22:00, long after sunset. Solar thermal loses on every count here:
+flat-plate and evacuated-tube collectors reach only the 60 to 90 °C that
+the feedwater heat pump already recovers from the press cooling return;
+concentrating collectors that could reach 180 °C need strong direct sun,
+which northern Thailand's haze and burning season do not give; and PV
+through a heat pump at a COP of ~3 delivers low-grade heat for ~1.2
+THB/kWh while also serving every other load. So the roof gets PV
+throughout and no solar thermal, and feedwater preheat stays with heat
+recovery.
+
+**The battery does not pay through the tariff.** Charging off-peak and
+discharging at peak earns ~1.1 THB/kWh after losses, ~70k THB a year for
+the 271 kWh stack, and spread over the 13-hour weekday peak window it can
+hold demand down by only ~20 kW (~30k THB a year) (estimates). Its case is
+ride-through and outage resilience, not peak shaving.
+
 ## Water supply
 
 The zero-discharge commitment below governs water *leaving* the plant. It