Here are some observations made over time of solar PV systems installed in the developing world, mostly in Africa, for the health sector and similar scale rural applications, as well as some thoughts for change to improve the viability of the technology's use in the health sector.
A) Installation standards
While some comprehension of good standards for equipment is understood and is often well documented and adhered to in supply tenders, installation standards are either neglected when services are procured, or not adequately monitored and evaluated during and after the completion of work. Some examples of common issues relating to installation weaknesses which lead to failure are as follows:
Array Wiring
• Use of non-UV resistant wiring for module and sub-array interconnects
• Use of non-UV resistant wiring for array to controller / load
• Improper sizing of wire gauge relative to length from array to controller / load (energy loss)
• Penetration of array to controller / load cable into buildings too often done any which way instead of through the side of the building on the opposite side of the direction of prevalent winds
• Too many times, I've seen a + and a - wire off of each module, and clumps of wires twisted and electrical taped together, and then to a main cable leading down to the array. Most of these systems are small (2-6 modules) and connections can be made in the array junction boxes. If larger, proper DC combiner boxes should be used.
• Strain relief is almost never seen. Zip ties are cheap.
• Silicone sealant is cheap and readily available, and should be used to weatherproof the entry of cables into junction boxes, even if grommets are supplied
• There should always be a foot or more of slack in array to controller / load cables, so that a 'U' shaped drop can be allowed for, in case water runs in along the wire. Without it, water will likely get to the controller and fry it.
Mounting structure for arrays
• Insufficient ventilation under modules so they run hot, deliver less power, and encapsulate degrades faster
• Mostly installed flat on zinc roofs, irrespective of orientation of building
• Should ideally be anodized aluminum but usually some local weld job with messy rust resistance paint which doesn't last long.
• Design and installation should not cause stress on module glass or framing - can lead to unnecessary breakage.
Breakers
• Most locally procurable off the shelf A/C electricity breakers are not appropriate for solar systems and power loss is often greater than 25%. Local installers don't know the difference and batteries will take a beating because they're not being charge as per the system design, or direct loads like fridges will under-perform.
Battery boxes
• These tend to be a joke. They need to be designed so that NOBODY on site can get into them for any reason. Central/regional/district technicians should be the only ones with access. Operators will connect their own loads whenever possible, even if it's just a cell phone.
Controllers
• Controller wiring should not be accessible to operators or this defeats the purpose of the battery box
• Spare fuses should always be kept on site next to the controller
Batteries
• If people can't maintain flooded batteries, consider larger capacity AGM sealed or even industrial forklift batteries (like the old Dekka maintenance saver)
• Build some passive solar thermal cooling into the housing design.
Grounding
• A copper ground rod, lug, braided copper cable, and some salt and charcoal can be purchased locally but you almost never see any kind of grounding on rural PV systems installed by projects
Housing
• Concrete blocks, zinc roof, with one or no windows = oven. Picking an existing building to install a PV system and fridge into is short sighted. Fridges and batteries are designed with certain operating conditions in mind, which affect not only their basic day to day performance, but also their longevity.
B) Supply issues
• Although I've never been a fan of the shorter lived solar PV module technologies (thin-film, amorphous etc), it seems that it would make better financial sense to procure all parts including modules for a 5 year lifecyle (match batteries), and invest the module savings into plug and play pre-integration of systems and training by the integrator for local companies contracted for installation of systems. MC-Connectors for the entire system and good quality wiring, as well as intelligent anodized aluminum mounts would probably make a substantial impact on the standardization of installations and the subsequent ability to maintain/repair/refurbish them.
• LCD displays on controllers tend to break. LED indicator lights on controllers rarely break. Nobody is reading the numbers. Technicians have multi-meters.
• Module junction boxes should come with at least 2 junction weatherproof box cable strain relief cable penetration grommets
C) The need for change
Buying equipment and dumping it on countries (local offices or local government) with a budget to install, train and maintain doesn't really work. Contract solar PV system integrators to procure everything, install, train, and maintain systems for a set timeframe like 5 years. They would only get paid for systems that are running and get income docked for downtime. Build GSM/GPRS live web based monitoring into the project cost and outsource monitoring to a third party. If integrators want to open a country office, they can; if they prefer to partner will a local firm and subcontract, they can. The bottom line is that the project is designed in such a way that unless those systems run for 5 years with upwards of 90+% uptime, they won't make a profit. Invariably, this will be building local capacity in the private sector to not only install but also maintain systems, and at the end of the 5-year contract, the trained local firm can can take the lead with periodic retraining and regular monitoring and evaluation by the funding agency. Ideally, this would be a set up as a BOOT (Build, Own, Operate and Transfer) program with gradual transfer over to the local company who would eventually structure a fee for service agreement with the government.
