What a survey of a corrugated shed roof has to measure before anyone quotes a kilowatt figure: purlin section and spacing, real sheet thickness, corrosion, existing services load, and who signs the report.
A factory in Gazipur wants 350 KWp on a corrugated shed. The first serious question is not which module or which inverter. It is what sits under that sheet, what is already hanging from it, and whether the frame can carry a new load for the next twenty five years without anybody having to think about it again.
This is the part of a rooftop project most often skipped, because it is the part that reduces the number of kilowatts a contractor can sell. A survey done properly sometimes comes back saying the roof takes 220 KWp, not 350. That answer is worth more than the difference.
An array does not sit on a roof. It sits on a sheet, which sits on purlins, which sit on rafters or trusses, which sit on columns, which sit on foundations. Every kilogramme you add travels the whole of that chain, and the chain is only as good as its weakest link. On most Bangladeshi industrial sheds the weakest link is not the truss. It is the purlin, or the fixing into the sheet.
There are also two quite different loads to consider. The gravity load of module plus rail plus clamp plus structure is modest and spread out. The uplift load is not. A shed roof in a nor'wester sees suction, and an array raised above the sheet gives the wind something to get under. At the roof edges, and especially at the corners, the suction coefficients are the highest anywhere on the building. That is why a sensible layout keeps a setback from the eaves and the gable, thins the modules at the corners, and puts more fixings there than in the middle of the field. The fixing that fails first is almost always a perimeter one, and it fails upward.
The design wind speed comes from the zone value in the Bangladesh National Building Code for that location, adjusted for terrain and exposure. A coastal site near Cox's Bazar and a sheltered site inside a built-up area of Narayanganj are not the same problem. A mounting supplier's claim that their system is rated to a certain speed refers to their own aluminium, not to your shed.
Purlin spacing sets the rail span, and the rail span decides which mounting system is possible at all. Cold-formed C and Z purlins on newer sheds sit at spacings of roughly a metre to a metre and a half. Older sheds framed with MS angle or pipe purlins can be far wider and far less predictable, and some have been extended by a previous owner with whatever section was in the yard that week.
Measure and record at several bays, not one:
The reason for measuring rather than reading a drawing is that the shed as built and the shed as drawn diverge. Purlins get cut for a duct. A run gets replaced after a fire. Somebody adds a mezzanine and hangs a hoist rail off the frame. The drawing, where it exists at all, records intent.
One purlin consideration gets missed more than any other: where the module clamps land. Every module datasheet specifies permitted clamping zones along the frame, and the mechanical load certificate only applies inside those zones. If the purlin spacing forces you to clamp outside the permitted range, you have voided the module warranty and the frame may deform under wind loading. Either the rail layout suits the purlins, or the purlins get supplemented. Do not solve it at the clamp.
Corrugated galvanised sheet sold here is described by a gauge number that often flatters the actual base metal thickness. The number that matters is the steel under the zinc, measured at a cut edge with a micrometre or with a coating-compensating gauge. It matters because the pull-out capacity of a self-drilling screw or a J-bolt is a direct function of that thickness, and because thin sheet distorts around a loaded fixing long before it tears.
Then map the corrosion. Walk the roof and photograph the places that always go first:
If the sheet has fewer years left in it than the array, say so and price the reroof into the project. Nobody wants to demount and remount a twenty five year asset in year four to change a roof covering. Replacing the sheet before the modules go on is cheap. Replacing it afterwards costs the reroof, plus the demount, plus the remount, plus the lost generation, and it is a conversation nobody enjoys.
The frame was designed for the sheet, the wind, and whatever allowance the original engineer made for services. Twenty years of operations use that allowance up quietly. A survey should list and locate every one of these, with an estimate of weight:
Then look down. Sheds in Bangladesh rarely fail at the truss. They fail at the column base plate, in a bay where the floor is washed down daily or where a drain has leaked for a decade, and the corrosion hides under a skirt of concrete. Break out a sample base where the floor conditions justify it.
Three different people can hold an opinion about your roof, and only one of them should be putting a stamp on it.
The mounting structure supplier can certify their own product. Their calculation covers the rail, the clamp and the bracket under a stated wind pressure. It says nothing about your purlins.
The solar salesman can tell you what fits geometrically. That is a layout exercise, and their commercial interest points in one direction: more modules. This is not a character judgement, it is simply how the incentive works, and it is exactly why the person who signs off the structure should not be the person whose commission depends on the answer.
The structural engineer takes the measured survey, models the existing frame under the added dead load and the revised wind case, and states in writing what the roof can carry, where, and with what strengthening if any. That document is what your insurer will ask for after a storm, what a lender's technical adviser will want during due diligence, and what protects you if a fixing lets go. A serious EPC contractor arranges it and prices it into the design rather than offering it as an extra you can decline to save money.
If you are writing this into a tender, specify the deliverable rather than the activity. A survey report that earns its fee contains:
A roof that fails the check is not the end of the project, it is the start of a different design. The usual options, roughly in order of cost:
Vvon has commissioned over 7 MWp in Bangladesh since 2017, including 1,503 KWp at Akij Agro Feed and 575 KWp at Ahad Jute Mills, and the structural survey has come before the layout on every one of them. It is the cheapest hour of engineering in the whole project, and the only one that cannot be redone later.