Net-metering pays you two different rates without saying so. This is how the credit works against an industrial bill in Bangladesh, how to size a plant to your daytime load rather than your roof, and what changes when your consumption profile moves.
Bangladesh imports a meaningful share of its primary energy and the industrial tariff for medium-voltage consumers has steadily climbed since 2022. For mills, hospitals and universities with a usable roof, an on-grid net-metering solar system is now the most reliable way to lock in electricity costs for the next 25 years. What decides whether a particular plant earns well is not the technology, which is settled, but the fit between the generation curve and the site's own load curve.
Under the Net Metering Guidelines 2025 (Power Division, administered by SREDA), eligible consumers, both single-phase and three-phase, can install rooftop solar of up to 100% of their sanctioned load (80% of transformer capacity for MV consumers) and feed surplus generation back into the grid. The utility nets imported and exported energy on the bill, and any surplus at the end of the financial year is settled at the prevailing BERC bulk-purchase rate. The customer keeps a single bi-directional meter and the same connection. No separate licence is required.
This is where most first proposals go wrong. A kWh your machinery consumes at the moment the panels produce it is a kWh you do not buy, so it is worth your full retail industrial tariff. A kWh exported because nothing on site was running is netted against a future import, and whatever remains unused at the end of the financial year is settled at the bulk-purchase rate, which is not the retail rate you pay.
Net-metering therefore pays two different prices, and the ratio between them is set by your own load pattern, not by the equipment. Two plants of identical capacity generate the same kWh per kWp and return different money. Ask any EPC contractor quoting you to show the expected split between self-consumed and exported energy, month by month, and to state which tariff each half was valued at. If they cannot produce that, the savings figure in the proposal is a guess.
Demand charges behave differently again. They follow your maximum demand, not your energy total, so solar reduces the energy line reliably and the demand line only when generation is strong at the moment your demand peaks. A mill whose peak is set by a morning compressor start on an overcast day sees the energy line fall and the demand line barely move.
Bangladesh's mid-latitude climate produces roughly 105 kWh per kWp per month on a well-oriented south-facing tilt, with monocrystalline modules from Tier-1 OEMs. A clean rule of thumb: every 1,000 sq ft of usable, shadow-free roof can host about 13 kWp of PV. The table below shows how that scales for typical Bangladesh facilities.
| Facility type | Usable roof | Plant size | Yearly generation | Yearly bill saving (BDT 10/kWh) |
|---|---|---|---|---|
| Garments factory | 60,000 sq ft | 780 kWp | 9,82,800 kWh | BDT 98,28,000 |
| Jute / textile mill | 45,000 sq ft | 585 kWp | 7,37,100 kWh | BDT 73,71,000 |
| Tertiary hospital | 30,000 sq ft | 390 kWp | 4,91,400 kWh | BDT 49,14,000 |
| University block | 20,000 sq ft | 260 kWp | 3,27,600 kWh | BDT 32,76,000 |
| Cold-storage shed | 100,000 sq ft | 1,300 kWp | 16,38,000 kWh | BDT 1,63,80,000 |
These are first-pass numbers. The real design must account for sanctioned load, transformer rating, roof orientation, shading, structural capacity (PSC vs corrugated metal), and the local utility's transformer headroom. Vvon's engineers run a full SLD plus structural and shadow analysis before quoting capacity.
Note which constraint usually binds. The regulatory ceiling is frequently below what either the roof or the load would justify, so if your sanctioned load sits under your connected load, a load revision may raise the approvable capacity and is worth pursuing before you design.
The roof gives you a ceiling. The load curve gives you the answer. Take twelve months of bills, get daytime consumption between roughly 9am and 4pm from your energy meter or a temporary logger, and compare it against the generation profile of the proposed array. Facility types behave very differently once you do.
A mosque with five prayer-time peaks, or an office block with an evening-weighted load, is the mirror image of a cold store: generation and consumption barely overlap, nearly everything is exported, and the return leans entirely on the settlement rate. Solar can still make sense there, but nobody should be shown a savings figure built on retail-tariff arithmetic.
On a typical industrial rooftop project in Bangladesh, the levelised cost of solar electricity is now well below the medium-voltage industrial tariff. The two flagship case studies Vvon engineered in the last three years illustrate the range:
| Case study | Capacity | Yearly generation | Yearly saving | ROI | CO₂ reduction |
|---|---|---|---|---|---|
| Akij Agro Feed Ltd. (Narayanganj) | 1,503 KWp | 18,96,660 kWh | BDT 2,01,04,500 | ≈ 3.5 years | 1,150 t / yr |
| Ahad Jute Mills Ltd. (Jashore) | 575 KWp | 7,31,500 kWh | BDT 77,54,000 | ≈ 4 years | 445 t / yr |
Both are continuous-process industrial sites with strong daytime load, which is why they sit at the good end of the range. Beyond payback, the residual life of a Tier-1 PV plant is 25 to 30 years, meaning roughly two decades of effectively free electricity once the capital is recovered. That is why net-metering has become the default capex decision for Bangladesh's larger industrial groups. Treat any payback figure, including these, as the output of assumptions: the three that move it most are your current tariff, the share of generation you consume on site, and the tariff path over the next decade.
A plant is a 25-year asset installed against a snapshot of how your business ran last year.
| Change at the site | Effect on net-metering value | Design response |
|---|---|---|
| Second or third shift added | New load is at night, so solar offsets a smaller share of a bigger bill | Be clear that solar addresses the daytime portion only |
| Line electrified or expanded | Daytime load rises, more output self-consumed at retail tariff | Recheck sanctioned load and transformer headroom |
| Line shut down or output cut | Exports rise, average value per kWh drops toward the settlement rate | Do not size to the ceiling where order books are volatile |
| Generator replaced by grid supply | Grid consumption rises, self-consumption improves | Confirm inverter behaviour while running on generator |
| Site relocates or the lease ends | The asset sits on a roof you no longer occupy | Match tenure to payback; price relocation into the contract |
| New block built next door | Permanent shading across part of the array | Design a shading margin; keep the layout able to be restrung |
The last row is the one that catches people. Shading creep is the most common quiet loss on Bangladeshi urban roofs: a neighbouring building goes up two storeys higher, or your own water tank is relocated, and one string underperforms for years while total generation still looks broadly right. String-level monitoring is what catches it.
Vvon Technologies Limited has commissioned over 7 MWp of industrial rooftop solar across Bangladesh, including the 1,503 KWp Akij Agro Feed plant in Narayanganj and the 575 KWp Ahad Jute Mills plant in Jashore. We are an EPC contractor working from Baridhara, Dhaka: we engineer, procure, build, commission and maintain complete plants rather than selling components. Equipment specified into our projects comes from Tier-1 makes including Jinko Solar, Canadian Solar, JA Solar, Sunways, Huawei, Growatt, Solis and Schneider Electric, chosen per project on yield, warranty and service reach inside Bangladesh.