A hospital cannot tolerate an interruption during a procedure, which rules out treating solar as a simple bill-reduction exercise. How the tiered critical-load architecture works, what the roof and the supply quality demand, and how to write the specification.
Hospitals and clinics operate 24 hours a day, 365 days a year, powering equipment that cannot tolerate even a momentary interruption: ventilators, operating theatre lights, ICU monitors, imaging systems and blood bank refrigerators. They also carry electricity bills that can reach BDT 20 to 50 lakh per month for a 200-bed hospital. Those two facts pull a solar project in different directions, and a design that only answers the second one is the wrong design for a hospital.
Bangladesh's healthcare sector faces a particular energy dilemma. Grid electricity is available in most urban areas but subject to voltage fluctuations and occasional outages. Diesel generators provide backup but at a cost of BDT 28 to 35 per kWh, approximately 2.5 to 3 times the grid tariff. A well-designed solar and battery system can reduce total energy costs by 35 to 50% while providing backup power that switches in under 20 milliseconds, faster and quieter than any diesel generator.
The point that matters commercially is the second one. If the hospital's generator currently runs for a meaningful number of hours each month, those are the most expensive units on the site, and displacing them is worth several times as much per kWh as displacing grid units. Ask for the generator running-hours log before anyone models a return. It is usually the most useful document in the building and it is almost never included in a feasibility request.
For hospitals, Vvon designs a tiered power architecture: solar and battery cover the critical load, meaning ICU, operating theatres, emergency and the blood bank, with automatic switchover in under 20 milliseconds; the grid covers the general load of wards, offices and canteen; and the existing diesel generator serves as a tertiary backup. This architecture minimises diesel running hours while ensuring zero interruption to life-critical equipment.
Making that real means someone has to draw the line between critical and general, circuit by circuit, on the existing distribution boards. In most Bangladeshi hospitals that separation is partial: theatre lighting is on the essential board but the theatre air handling is not, or the blood bank refrigerator shares a circuit with a corridor. Expect the survey to find this, expect some reworking of the boards to be part of the project, and get it into the scope in writing. A solar contractor who quotes a battery without having traced the essential distribution has quoted for a different building.
| Hospital size | Beds | Monthly consumption | Solar plant | Battery storage | Monthly saving |
|---|---|---|---|---|---|
| Small clinic | 20 beds | 15,000 kWh | 50 kWp | 50 kWh | BDT 1,89,000 |
| District hospital | 100 beds | 60,000 kWh | 200 kWp | 150 kWh | BDT 7,56,000 |
| Tertiary hospital | 300 beds | 180,000 kWh | 500 kWp | 400 kWh | BDT 22,68,000 |
| Specialty hospital | 500 beds | 300,000 kWh | 800 kWp | 600 kWh | BDT 37,80,000 |
Note that the battery figures in that table are a fraction of daily consumption, not a day of autonomy. They are sized to hold a defined critical schedule through an outage and a switchover, not to run the hospital off-grid overnight. Any proposal that implies otherwise is either far larger than the price suggests or has not defined the load it is protecting.
Every hospital that already runs safely has protection in place before any solar contractor arrives: UPS units on ventilators, monitors, theatre equipment and the servers, and a generator behind them. A solar and battery installation does not replace that, and an owner should be suspicious of anyone who implies it does. What it does is sit upstream, so the battery carries the essential board through the outage and the UPS units downstream stop being asked to work at all.
The practical result is that the existing UPS runtime becomes reserve rather than the front line, and the generator starts far less often. That is where the diesel saving in a hospital comes from, and it is worth stating in the specification as an outcome to be demonstrated at commissioning: a witnessed test in which the grid is dropped, the critical board is held, and the generator does not start. Ask for that test. It is the only way to know the separation was done properly, and it takes an afternoon.
Medical equipment is less tolerant of a poor supply than industrial plant, and hospitals have circuits with specific requirements that a general-purpose solar design will not know about. Operating theatres use isolated power supply systems with insulation monitoring, and those are not a place to tie in an inverter. Imaging equipment draws heavy, brief inrush during exposure, which sets the fault level and the voltage regulation the supply has to hold at the point of connection. Laboratory and imaging equipment also frequently comes with manufacturer conditions on supply quality, and breaching them can affect a service contract.
Hospital roofs are usually RCC slab, which is structurally the most forgiving surface to work with and the least forgiving to get wrong. The slab will carry the weight, so the design questions move to wind uplift on a ballasted or fixed system, and to waterproofing. A leak over a ward or a theatre is a clinical problem, so the penetration and sealing detail deserves more scrutiny here than on any factory job.
The other constraint is that hospital roofs are busy and hospitals grow. Plant rooms, lift machine rooms, water tanks, air handling units and their service access all occupy roof that looks clear from above, and the array layout has to leave working room around them rather than boxing them in. Ask the administration one question early: is there any intention to build another floor? Fitting a 25-year asset onto a slab that is scheduled for vertical extension in year four is a costly sequencing error and a common one.
The connection application goes to DPDC or DESCO for hospitals in Dhaka city, or to BPDB, NESCO or BREB elsewhere, on the SREDA-format application with the single-line diagram, equipment datasheets carrying IEC certification numbers, roof plan, a recent bill and the institution's documents. The plant is capped at 100% of sanctioned load, or 80% of transformer capacity for medium-voltage consumers. Expect a technical committee review and site inspection before the No Objection Certificate is issued.
Public hospitals procure through e-GP, and that changes what the specification has to do. Where evaluation is driven by price, the tender document is the only place where quality can be defended, so it needs to state the guaranteed performance ratio and the remedy for missing it, the module and inverter warranty terms, the required critical-load separation and switchover time, the surge protection and earthing requirements, the monitoring and reporting outputs, and the O and M period with response times. Left unstated, each of those becomes the first thing a low bid removes.