Solar Energy Audit for Thai Factories: Site Survey & Feasibility Guide

Factory A installed 800 kWp from a quick estimate, but actual load profile peaked at only 500 kW on weekdays and 200 kW on weekends. Excess power exported to grid earned nothing (no Net Metering). System oversized by 60%, ROI stretched from 4.5 to 9 years.

Factory B installed 200 kWp because the roof area "looked sufficient" but didn't check shading from chimneys and cooling towers — 25% shading loss meant actual production was only 150 kWp, falling below EPC performance guarantee.

You need at least 12 months of bills to see full-year usage patterns — summer (Mar-May) has highest consumption due to cooling load, while monsoon (Jun-Oct) reduces solar production by 15-25% from cloud cover. Extract: units consumed (kWh), demand (kW), power factor, Ft adjustment, and demand charge split by On-Peak/Off-Peak. Smart Meter 15-min interval data is even better for precise system sizing.

The load profile reveals hourly consumption patterns. Key points: (1) Base load (24/7 usage like industrial refrigeration, ventilation) (2) Peak demand (main machinery hours, typically 08:00-17:00) (3) Correlation between peak demand and solar production window (09:00-15:00). If peak aligns with solar window = high self-consumption = better ROI. If peak is evening/night = consider battery storage or adjusted design.

On-site roof survey checks 5 things: (1) Usable area (minus HVAC units, drainage, maintenance walkways) — typically 60-75% of total roof area. (2) Orientation — south-facing is optimal for Thailand; east/west produces 10-15% less. (3) Tilt — flat roofs need 10-15° tilt structures. (4) Shading from obstructions — chimneys, cooling towers, adjacent buildings, trees — measure shadow angles throughout the day. (5) Structural condition — roof must support additional 12-18 kg/m² (panels + mounting). Roofs >15 years old may need reinforcement.

Solar irradiance (sunlight reaching the site) is the primary yield calculation variable. Start with Solargis, NASA POWER, or Meteonorm data, but cross-check with actual conditions: industrial zones with dust/pollution may have 3-8% lower irradiance than satellite data. Coastal areas (Chonburi, Rayong) may see 2-5% reduction from salt haze. Average GHI across Thailand is 1,600-1,800 kWh/m²/yr, but site-specific values can differ by up to 10% due to microclimate.

Often overlooked but critical: (1) Transformer — check if remaining capacity can absorb solar output. A 1,000 kVA transformer at 800 kVA normal load leaves only 200 kVA headroom for solar. Exceeding capacity means replacing the transformer (THB 500K-2M depending on size). (2) Main Panel / MDB — must have space for new breakers and busbar rating must support the addition. (3) Grid connection — must apply to PEA/MEA for grid interconnection approval. Systems >1 MWp may require a Power Plant License.

With all data collected, this step "test-designs" the system: select system size (kWp) from load profile + roof area + grid capacity constraints → choose panels (N-type TOPCon or PERC) + inverter (string or central) → input everything into PVsyst simulation → get yield estimate (kWh/kWp/yr), performance ratio (PR), self-consumption ratio, specific yield. Standard PVsyst reports show: monthly/annual production forecast, loss diagram (shading, temperature, wiring, inverter), bankable P50/P90 yield estimates.

The Feasibility Report is the final Energy Audit deliverable: (1) Site survey summary (2) Recommended system (size, panels, inverter, layout) (3) Yield forecast + PR estimate (4) Financial Model — CAPEX, annual savings, payback period, IRR, NPV, LCOE (5) Sensitivity analysis (tariff ±10%, degradation rate, discount rate) (6) Risk + mitigation (grid limitation, shading, structural) (7) Implementation timeline + milestones. This document is the "decision document" that CEO/CFO uses for budget approval and banks use for loan approval.

12-15 months of recent electricity bills (PEA/MEA)

Smart Meter 15-min interval data (if available)

Building floor plans / roof structural drawings

Factory operating license (Ror Ngor 4)

Transformer data — size (kVA), brand, age

Single Line Diagram (SLD) of factory electrical system

Roof photos from multiple angles (overview + potential shading spots)

Google Earth / Maps satellite image with roof boundaries marked

Factory operating schedule (shifts, holidays, seasonal patterns)

List of major equipment + power rating (kW) for each

Future expansion plans (if any — adding machinery/buildings?)

Preliminary budget + constraints (EPC vs PPA preference)

Yield estimate is the annual energy production per 1 kWp. For Thailand, realistic values are 1,300-1,500 kWh/kWp/yr (depending on location, tilt, shading). If an EPC quotes yield >1,500 — be skeptical; they may be inflating numbers for better ROI optics. If <1,200 — may be overly conservative or have unresolved design issues (shading, orientation). When comparing EPCs, ensure same basis: same P50 vs P90, same degradation rate assumption.

Performance Ratio (PR) = actual output / theoretical output. New systems in Thailand should achieve PR 77-83%. If an EPC uses PR <75%, it signals abnormally high losses — possibly unresolved shading, undersized inverters, or high wiring losses. If PR >85%, it's likely overly optimistic, not accounting for Thailand's temperature losses (the primary loss factor). Good PR for Thai rooftop systems: 78-82% (PERC), 80-84% (N-type TOPCon/HJT).

The ROI and payback period in EPC proposals depend on 3 key assumptions you must verify: (1) Discount rate — should use 6-8% (real cost of capital), not 3% which inflates NPV. (2) FT escalation — Thai electricity rates tend to rise 3-5%/yr, but some EPCs use 7-10% to inflate savings. (3) Degradation rate — should use 0.4-0.6%/yr (Tier 1). If EPC uses 0.25%, it's overly optimistic. Request all 3 assumptions for side-by-side comparison across EPCs.

Monthly bills only tell you "how much per month" but not "when" — a factory using 100,000 kWh/month may peak in the morning (aligns with solar window) or evening/night (solar can't help). Sizing from average bills without load profile analysis → self-consumption may be 20-40% lower than expected → ROI collapses.

Shadows from chimneys, cooling towers, large trees, or even neighboring buildings under construction can reduce output by 10-30%. Especially critical with series-connected panels (string configuration) — shade on just 1-2 cells can reduce entire string output. Use proper tools (Suneye, Solmetric, or drone + 3D model) to analyze shading across all 12 months, not just the survey day.

The most expensive mistake: installing a 500 kWp solar system without checking that the existing 800 kVA transformer is already loaded at 750 kVA. When solar generates power, reverse current flows into the transformer causing overload protection to trip — system shuts down during peak production. Fix requires a new transformer (THB 1-2M) or limiting solar output to 50 kVA (wasting 90% of the system). Check before sizing — not after installation.

Systems >500 kWp — high investment (>THB 15M), significant risk from mis-sizing. IE review costs approximately THB 80,000-200,000, well worth the insurance against costly mistakes.

PPA contracts >15 years — if you're not buying but signing a 15-25 year PPA, total payments may reach THB 50-100M. IE verifies that yield estimates, escalation clauses, and degradation guarantees in the contract are realistic.

No in-house electrical engineer — factories without an in-house engineer to review should hire an IE as a "second pair of eyes," especially for grid interconnection and transformer capacity issues that require specialized expertise.