SOLAR-FIRST · PEAK SHAVING

Cut factory electricity 30-50% with Solar Rooftop + BESS Peak Shaving + GridMind AI

Factories pay two bills people overlook — energy (kWh) and a Demand Charge (kW) set by the single highest 15-minute peak of the month. LEONICS integrates Solar Rooftop + BESS + Modular PCS + GridMind AI: Solar clips On-Peak energy, BESS shaves the demand peak to cut Demand Charge 15-25%, and GridMind AI forecasts load and irradiance to dispatch the battery optimally — 30-50% lower total bills, 4-6 year payback, with CBAM-ready carbon reporting.

30-50%
Total bill cut
15-25%
Demand Charge cut
4-6 yr
Payback
12-15%
Project IRR

Design figures for a system sized to this use case. What a given site actually gets depends on its load shape, the PV and battery sizing and its tariff.

The Challenge

A factory's Time-of-Use (TOU) bill has two cost drivers that solar alone can't fix — plus a looming 2026 compliance deadline. Solar touches the energy charge; only BESS + AI cuts the Demand Charge.

  • Energy charge — On-Peak ~THB 4.10–4.33/kWh (Mon–Fri 09:00–22:00), nearly 2× the ~THB 2.59–2.64 Off-Peak rate. ERC base rates, before Ft and VAT.
  • Demand Charge — THB 132.93/kW/month at 12–24 kV, 210.00 below 12 kV — billed on the single highest 15-minute peak of the entire month.
  • On-Peak concentration — Production load piles into the most expensive window of the day.
  • Momentary spikes — One motor start or chiller surge for a few minutes inflates the Demand Charge for all 30 days.
  • CBAM 2026 — EU exporters of steel, aluminium and cement must prove verifiable Scope 2 reduction. The definitive period began 1 Jan 2026; certificates go on sale Feb 2027.

How LEONICS Solves It

01

Solar Rooftop 100 kWp – 5 MWp

Self-consumption on the factory roof — directly clips On-Peak energy (daytime load lines up with generation).

02

BESS (LiFePO4) + Modular PCS 0.3–3.2 MW

Charges Off-Peak or from surplus solar, then discharges for peak shaving — cutting the 15-min Demand Charge peak 15–25% without touching production.

03

GridMind AI

Forecasts solar and load up to 4 hours ahead, then recomputes the grid-import cap ("shave line") every 15 minutes — discharging into the morning peak when midday surplus is forecast to refill, and holding one flat cap across 09:00–22:00 to protect Maximum Demand.

04

Energy Monitoring + Carbon Report

Meters kWh, peaks, SOC and PV yield — an auditable carbon report ready for CBAM, T-VER, ESG.

Who makes it, who installs it

The core equipment is made by LEONICS Co., Ltd. Turnkey design and installation (EPC) of factory solar + battery systems in Thailand is carried out by Microgrid Technology Co., Ltd. (MGT), a LEONICS group company.

  • LEONICS — manufacturer — Designs and manufactures the Modular PCS, the C&I ESS battery system and the GridMind AI platform that dispatches the battery. Building power electronics since 1991, at its factory in Wellgrow Industrial Estate, Chachoengsao.
  • Microgrid Technology (MGT) — installer (EPC) — The LEONICS group's EPC contractor — design, procurement and installation of solar rooftop, solar farm and solar + BESS systems in Thailand. On MEA's People's Solar Vendor List (announcement No. 4/2569 of 15 October 2026, entry 00090).
  • MGT factory projects — Kerry Flour Mills 3.42 MWp (two signed Final Acceptance Certificates, 2024), Kaneka (Thailand) 517.92 kWp (2021) and Inpac Pharma 418 kWp (2023) — details under Proven Installations below.
  • Solar + BESS delivered by LEONICS in Thailand (commercial sites) — Designed and installed by LEONICS itself: GEMS at a Bangchak service station, Bangkok (2018 — 280.9 kW PV + 1 MWh LG lithium battery with AI forecasting) and the PTT service-station microgrid, Pathum Thani (2015 — 520 kW PV + 3 MWh battery). Installation in Thailand today is carried out by MGT.
  • Contact — Tel +66 2-746-9500 · MGT mobile 080-042-4469 · LINE microgrid-tech · microgrid-tech.com
Factory solar + storage — four layers that cooperate

Rooftop solar covers the daytime load, the BESS clips the 15-minute peak that sets the demand charge, GridMind AI decides when to dispatch, and metering closes the loop.

Factory solar and storage architectureA 100 kWp to 5 MWp rooftop array and a LiFePO4 battery connect to the DC side of an N+1 Modular PCS. Its AC output meets a bus shared with the utility supply, metered on the TOU tariff, and feeds the factory load. GridMind AI forecasts and dispatches to clip the 15-minute peak that sets the demand charge.Solar Rooftop100 kWp – 5 MWpBatteryLiFePO4 · peak clippingModular PCS + BESSPCS · N+1GridTOU tariffMeterkWh · peaks · SOCFactory loadlines · chillers · air compressorsGridMind AIforecast → dispatch → demand limitDCAC1DC→ACThe battery clips the 15-minute peak that sets the demand chargeDCAC

Schematic. Ratings shown are the configurable range, not one site; demand-charge reduction depends on the load shape.

Rule-based EMS vs GridMind AI — same hardware, one full year

On the identical site, GridMind AI extracts 55% more energy from the same battery and buys 149 MWh/year less peak-rate power — while holding Maximum Demand equal. Forecast-driven dispatch simply uses the asset you already paid for, better.

Rule-based EMS versus GridMind AI — annual outcomes on the same site On the same battery hardware, GridMind AI delivers 55% more battery energy per year and imports 13% less peak-rate grid energy than a tuned rule-based EMS, with Maximum Demand held equal. Battery energy used / year 273 Rule-based EMS 422 GridMind AI +55% MWh more value from the same battery Peak-rate grid imports / year 1,129 Rule-based EMS 980 GridMind AI −13% MWh −149 MWh of expensive peak energy Same site · 300 kWp PV + 750 kW / 1,566 kWh BESS · Maximum Demand held equal (687 vs 684 kW)

Source: full-year 15-minute Solar Insight Pro simulation of the same site — rule-based EMS (fixed peak-shaving target) vs GridMind AI (forecast-driven), at measured forecast-error levels.

Key Benefits

  • ✓30-50% lower total bill — Solar clips On-Peak energy, BESS cuts the Demand Charge, and AI optimises TOU, all at once
  • ✓15-25% lower Demand Charge — peak shaving trims the 15-minute peak that sets demand for the whole month, which Solar alone cannot do
  • ✓GridMind AI forecasts load + irradiance — dispatches the battery onto the real peak and auto-avoids On-Peak, no operator needed
  • ✓Forecast-driven, not rule-based — a conventional EMS hoards the battery because it has no lookahead; GridMind AI predicts ahead and delivers +55% more battery energy than a tuned rule-based EMS in a full-year simulation run at measured forecast-error levels.
  • ✓High solar self-consumption — daytime production load coincides with sunlight, so nearly all generation is used on-site, not exported for pennies
  • ✓Modular PCS 0.3-3.2 MW, N+1 — containerised and install-ready, scales by phase, and a single module fault never stops the line
  • ✓Auditable, CBAM-ready carbon report — proves Scope 2 reduction for EU exporters; supports CBAM, T-VER and ESG disclosure
  • ✓4-6 year payback, 12-15% IRR at current Thai industrial TOU rates and Demand Charges
  • ✓UPS / ride-through function — rides through voltage sags and momentary dips that trip lines and PLCs and scrap a shift's output

Best For

Food & beverage plants (heavy chiller / cold-storage load)Automotive & parts manufacturers (EU exporters facing CBAM)Textile & dyeing millsSteel / aluminium / metal casting worksPlastics / injection moulding (high motor-start peaks)Cold storage / frozen warehousingChemical & fertiliser plantsIndustrial estates / multi-tenant factories

Technical Specifications

Solar Rooftop (typical)
100 kWp - 5 MWp · Mono PERC / bifacial, Tier-1 module
BESS
LiFePO4 · 250 kWh - 4 MWh · 6,000+ cycles @ 80% DOD
Modular PCS
0.3-3.2 MW · containerized · N+1 parallel
Core functions
Peak Shaving · TOU Arbitrage · Demand Limiting · Self-consumption
GridMind AI
Load + PV forecast · automated dispatch · demand-cap enforcement
Demand Charge reduction
15-25% (from the 15-min peak)
Grid interface
Grid-tied / Grid-tied with backup · anti-islanding
Grid code
UL1741SA · IEEE1547 · IEC 61727 · MEA/PEA interconnection
Metering & reporting
Modbus TCP / SNMP · kWh + kW peak + SOC · carbon report
Carbon reporting
CBAM · T-VER · ESG ready
Forecast horizon (GridMind AI)
Up to 4 h · recomputed every 15 min
Forecast error (measured)
9–13% wMAPE at live sites
Demand charge control
One flat import cap across 09:00–22:00

Proven Installations

Real installations by LEONICS and Microgrid Technology (a LEONICS group company)

View all projects

Frequently Asked Questions

Who installs the solar + BESS system at the factory?
LEONICS Co., Ltd. manufactures the core equipment — the Modular PCS, the C&I ESS battery system and GridMind AI. Turnkey design and installation (EPC) in Thailand is carried out by Microgrid Technology Co., Ltd. (MGT), a LEONICS group company whose factory rooftop projects include Kerry Flour Mills (3.42 MWp), Kaneka (Thailand) (517.92 kWp) and Inpac Pharma (418 kWp). Before that, LEONICS designed and installed solar + BESS in Thailand itself — GEMS at a Bangchak service station (2018, 280.9 kW PV + 1 MWh battery) and the PTT service-station microgrid in Pathum Thani (2015, 520 kW PV + 3 MWh battery).
How do Solar Rooftop and BESS each cut a factory's bill, and why are both needed?
A factory bill has two parts. Solar Rooftop cuts the energy charge (kWh) — self-generating by day instead of buying On-Peak power at ~THB 4.18-4.33/kWh. But Solar does not reduce the Demand Charge, because demand is billed on the highest 15-minute peak, which often occurs under cloud cover or at night when solar can't help. BESS is what does peak shaving — discharging to clip that peak and cut the Demand Charge 15-25%. Together they reach 30-50% savings; Solar alone typically saves only 20-30% and never touches the Demand Charge.
What is the Demand Charge, and how much can peak shaving really cut it?
The Demand Charge is billed on your highest 15-minute power peak in the month, times a rate of ~THB 132-210/kW/month (depending on voltage level). A single peak — say a midday chiller starting alongside large motors — sets the charge for all 30 days. BESS + GridMind AI detect an emerging peak and instantly discharge so the demand drawn from the grid never crosses the cap — typically a 15-25% Demand Charge cut. Factories with spiky loads (motor starts, batch processes) save even more.
What's the payback period and IRR for a typical factory?
Typically 4-6 year payback at 12-15% IRR under current Thai industrial TOU rates. Payback accelerates with: (1) a high On-Peak load share (2) a high per-kW Demand Charge (sub-22kV factories pay ~THB 210/kW) (3) spiky peaks that peak shaving can trim deeply (4) BOI incentives or accelerated depreciation. Projects where Solar+BESS are sized to the real load profile often hit near 4-year payback, while Solar-only stretches to 6-8 years because it never touches the Demand Charge.
How does the system help with CBAM and carbon reporting for exports?
EU CBAM entered its definitive period on 1 January 2026 for steel, aluminium, cement, fertiliser, electricity and hydrogen. Certificates go on sale February 2027 and the first surrender for 2026 imports falls due 30 September 2027; importers under 50 tonnes a year are exempt. Exporters must report embedded emissions including Scope 2 (purchased electricity). Solar Rooftop + BESS reduce grid imports carrying a high emission factor = verifiable Scope 2 reduction. The LEONICS energy monitoring system produces an auditable report: RE kWh, avoided Scope 2, and RE ratio — in a format ready to submit for CBAM, T-VER and ESG, with no manual back-calculation.
Will installation disrupt production or require a factory shutdown?
No production shutdown is required. Solar Rooftop is installed on the roof while the plant runs normally; the grid tie-in happens during a short, pre-scheduled maintenance window. BESS + PCS are containerised, placed beside the building and wired into the main distribution board. The system is grid-tied with backup — it parallels the utility under normal conditions without disturbing the load, with standards-compliant anti-islanding. The UPS / ride-through function also guards against voltage sags that trip PLCs and servos, cutting in-shift scrap. We commission in phases so the line keeps running.
How is GridMind AI different from a fixed charge/discharge schedule?
A fixed schedule (e.g. charge at midnight, discharge at 2pm) works only when load and sunlight are constant — but real factory load swings with the production plan and solar varies with cloud cover. GridMind AI forecasts ahead — both load and solar output per interval — then computes how much energy to hold back to clip the peak that's actually about to happen and to discharge across On-Peak for maximum value. The result is a deeper Demand Charge cut, with no peak overshoot from the battery emptying at the wrong moment. It's the same forecasting technology LEONICS runs in production solar/load forecasting at its Bangkok site.

Interested in this solution?

LEONICS designs custom solutions with ROI/Payback estimates before installation.