Power ideas for off grid hospital in Sierra Leone, West Africa.

cruiser guy

We are planning the power system for an off grid hospital in Sierra Leone West Africa. Forget any ideas of the term "hospital" you have and think of a large doctors office that will eventually have a surgical suite at some point in the future. For now it is only outpatient care.

Currently we have 1080 watts of panel on the roof powering the compound water system which is a Grundfos SQFlex pump.

The hospital will be constructed/rehabilitated in stages so we'd like to be able to start small and add as needs require rather than jump in and get all that will be needed when it's all said and done right now. A generator will be expected at some point, maybe even now.

Power needs at this point are quite minimal, a few lights here and there (fluorescent already installed prior to my arrival) and power for a few medical items like a microscope. Country power is 230vAC 50Hz but as we expect some of the medical equipment to be donated from the U.S. we will also require a 110vAC 60Hz system. Both have been wired already as separate panels.

I've e-mailed the Grundfos rep. and he says it is possible to both charge batteries and run the pump from the same solar panel array. I'm still awaiting a wiring diagram but for now lets go with what I have been told.

What are the ideas out there from those who have been into this much longer than I?

westbranch

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

CG, what is your best 'worst scenario' guess as to the total 230 and 120V loads.
Since you will most likely be using sensitive electronic devices a TSW inverter (s) would be called for.
based on the current array size you will probably want to go for a top line CC ie Classic from Midnite or similar, size TBD.
Battery bank TBD after loads determined

Cariboocoot

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

Starting small and adding as needed is one of the most difficult things there is to do with off-grid power. You need a plan. If you can say "we need 'X' kW hours now, 'Y' kW hours next year, and 'Z' kW hours in two years" then you could do it. Otherwise the expansion will either be a nightmare or reconfiguring/rewiring or just plain adding entirely separate systems to handle each new batch of loads. The latter is preferable in some ways as it provides redundancy which can be valuable for keeping critical things going in the event something fails.

Having to have two separate frequency/Voltage systems is bound to be trouble as soon as someone gets their hands on the wiring and doesn't understand the difference. Isolate the systems as much as possible. Labels, conduit, and locked access if you can. The last thing you want is someone plugging an expensive piece of medical equipment into the "wrong power" and burning it up. I trust you have the variety of outlets/plugs so that the 230/50 will use "British" connections and the 120/60 will use "American" connections? Even so, one enterprising idiot with a screwdriver ...

Now, what can you get from the existing 1080 Watt array? First issue would be how much of its production goes to water pumping. That will determine how much power you'd have "left over" for recharging batteries. Unfortunately that is going to vary with water usage. If you put priority on the pump, you may find you don't have enough power to recharge the batteries on certain days. As such you'll have to play it conservatively, and monitor it closely.

One way of looking at it is that the array could provide the peak current for 690 Amp hours @ 12 Volts. But this isn't the way to look at it because of the power going to pumping. So try this: 1080 Watts in a 5 hour day can produce about 2.8 kW hours AC (this is also a problem because the pumping is DC, but going with the lower value is safer). Now how does that compare to the projected light usage?

It will also be important to know the array specs, as the pump probably operates at a higher Voltage than would normally be used for charging batteries. This means it would be necessary to use an MPPT type controller, which adds expense and a certain amount of complexity.

If you could use the panels with a PWM controller than has LVD to control the lighting power, then you could have it shut off power before the batteries get too low. This can be done with the MPPT as well, but again it's more complex.

The obvious simplification would be to run the pump from batteries too. That way they become an integral part of the DC system and the panels just have to keep the batteries charged.

Just a few thoughts. I'm sure others will have some as well.

BB.

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

You usually want to avoid paralleling two MPPT type loads (pump and solar charge controller) on the same array... They will just confuse each other.

Either make a transfer switch (Manual or automatic--Recommend you switch both positive and negative legs so you avoid DC Grounding issues--Perhaps the transfer switch can be controlled by the tank pressure/level sensor).

Or, if this is pump to a cistern type setup--Get a MPPT controller with programmable "in float" status output (Outback and Midnite have? perhaps others?). Enable pumping when the batteries are reasonably full and you have "power to burn".

Otherwise, as Marc says--Plan on DC power to the pump (gives you 24x7 pumping ability plus automatic backup with a genset).

-Bill

cruiser guy

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

Power use guesstimates. It really all depends on what is donated. Questions like, "Will there be an x-ray machine?" are answered by "If someone donates one!". I know that's not great for design but there really is no way to know what might "fall off the truck" next year or in five years. I also can't say how long until the hospital is completely built/renovated. That too depends on donations. Maybe two years, maybe 20 years, no one really knows!

When adding panels or batteries we can simply keep adding to the existing bank can't we? So long as the charge controllers all have their own dedicated array if they all feed into one battery bank I'm not going to harm anything am I? I would imagine that the charge controllers likely should all be able to communicate with each other so equalization and so on all happen at the same time with each charge controller but other than that there's no problem. We'll likely have an automatic start for a diesel genset and monitor how often the genset runs to know when it's time to add panels.

Lights are all 230vAC fluorescent lights which use a starter, not great for energy conservation but that's what we've got. I'm going to propose that we retrofit to LED type lamps for lights which are on most of the time. How many lights are on? Again, tough to say.

The two separate power systems are separate at the panels. All 230v 50Hz sockets are the British standard and all 110v 60Hz sockets are the North American standard. I propose that the hospital be a "power bar free" zone! Power bars sold here accept ALL plug types including the American ones. Yes, I can see some equipment being fried when plugged into the wrong source but if we are strict on the power bar rule we should be reasonably OK. Wiring, unfortunately, is not in a conduit so some clever guy in the ceiling could screw things up badly.

For inverters I was thinking of the Outback power off-grid inverters both the 230v and the 110v ones. These start at about 3kW but can be added to simply by adding another of the same type inverter. As I understand it, they will synchronize their wave form and work as one inverter. This should make additions in the future a little easier (assuming Outback continues to manufacture these inverters or ones which are compatible).

For batteries, a 12v 200Ah Deka Gel cell deep cycle battery goes for about $750 ea. here. We've been using 12v 200Ah lead acid truck batteries in our house as they are cheaper at $150 or so each and since we have no kids to play on them the acid is not a concern. The hospital equipment will also be in a locked, ventilated room which is only accessible from outside. I was going to simply use standard truck batteries for now in series till we reach our desired battery bank voltage, probably 48v. In the future when all is finished then when the battery bank goes out we could change over to a more robust deep cycle battery at that time as we'd have a good idea of loads by then.

The pump doesn't run much more than an hour or so a day with the current occupancy on the compound. Currently there is us as a couple and a family with four children connected to the water system. We have a guest house that will be used intermittently as well as the hospital where we will need to keep a lid on water buckets walking out the door once the water is turned on but neither of these is up and running yet. The pump will often run as soon as there is sufficient sun to make up for water used in the previous evening or that morning, leaving the prime solar harvesting time of noon/early afternoon available. The pump is also connected and can be used off a generator if needed or when a generator is running.

The current solar array is 12 Solar World 90 watt panels all connected in a series string. This is done as the Grundfos SQFlex works most efficiently at higher voltages. There is nothing stopping us from changing the array configuration to two groups of six or some other configuration if that works better.

Grundfos has said that it is possible to run a charge controller with their system. I am awaiting a wiring diagram of exactly what they propose. It may be running the pump off the battery bank, I don't know for sure.

I was thinking of the tank float switch as a means to switch between the Grundfos control and the regular charge controller. Does anyone know if the tank level contacts in the Grundfos CU200 can be linked through a relay of some kind? I know the tank float switch opens on demand as opposed to what one would normally think and close on demand.

To keep things working in the future I'm hoping I can find and train a smart kid to be the maintenance guy. So far no luck in that regard. Plenty of kids but there is a very low literacy rate here.

inetdog

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

cruiser guy wrote: »

Lights are all 230vAC fluorescent lights which use a starter, not great for energy conservation but that's what we've got. I'm going to propose that we retrofit to LED type lamps for lights which are on most of the time. How many lights are on? Again, tough to say.

Actually, T5 type fluorescent lights are slightly more efficient than LEDs at the moment, although that may change.
If you have older T40 type tubes and ballasts, you may get a better return from retrofitting "energy saving" tubes that can use the same ballasts.
What the LED should give you though is much longer lifetime. But there is still active electronics in the LED package and overdriven LEDs can lose light output, so that is not a guarantee at this time.
What I am trying to say is that if you have fluorescent already, the ROI on retrofitting to LEDs will not be nearly as good as other potential uses for the money.
Replace incandescents wherever you can though.

PS: The starter itself does not use much energy, so you can leave that out of the equation. It is the tube and ballast combination, as well as the design of the fixture to get the most light out of the tubes, that really makes a difference.

Cariboocoot

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

When adding panels or batteries we can simply keep adding to the existing bank can't we? So long as the charge controllers all have their own dedicated array if they all feed into one battery bank I'm not going to harm anything am I? I would imagine that the charge controllers likely should all be able to communicate with each other so equalization and so on all happen at the same time with each charge controller but other than that there's no problem. We'll likely have an automatic start for a diesel genset and monitor how often the genset runs to know when it's time to add panels.

Okay, this is something you definitely do not want to do: randomly add on batteries/panels as they appear. For one thing, mixing new batteries with old will cause the new batteries to immediately lose some capacity as they try to charge the old. How bad this will be depends on what state the old batteries are in when the new ones are added. It can be like connecting batteries of vastly different capacities. What's more, if you increase the size of the battery bank you need to increase the size of the PV array to keep up with the increased charging demands. Multiple controllers on one bank is usually not a problem, even if they are different makes/models. Sometimes they can't be programmed alike, so one will end up contributing to Bulk and not doing much of anything else (which is usually not a problem).

I would avoid gel batteries as they are the pickiest for correct charge parameters. Flooded cells can take quite a bit of "abuse" before they are useless.

The Grundfos will run on 30 to 90 VDC, which means the 'logical' all-battery system would be 48. This has advantages for high-power systems, but quite a draw back for low-power systems; there aren't any low-power 48 Volt inverters. For just running lights, as an example, you really don't need to feed a 3kW inverter; 300 Watts would do.

Outback inverters would be great here, as they have a well-deserved reputation for standing up in harsh environments and as you know come in both European and North American versions. Of course if someone gets confused about which is which there ... big trouble.

You might want to think about a small 12 Volt system just for the lights to begin with. A 200 Amp hour truck battery and a MS 300 Watts inverter (be sure to get the 230 Volt version) would keep a standard "40 Watt" tube light going for 30 hours, or three of them for 10 hours. The only trouble here is using the PV to recharge due to Voltage differences.

I don't know the specs on the SolarWorld panels, but I can't see how 12 in series would fall in to the Grundfos's Voltage range. That would have to be at least 210 Volts which is far higher than the pump needs/should have. If they are standard 17.5 Vmp the array would be three parallel strings of four in series for Vmp of 70 (standard for a 48 Volt system). It may be necessary/possible to reconfigure only for battery charging, and revert to higher Voltage for pumping. But it's complicated.

westbranch

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

Just a thought, as there is a fair possibility that someone could mix up the 110 and 230 V systems when additions/modifications are done, perhaps you might consider using totally different types and COLOURED wires for the 2 systems and also different makes of inverters, also painted the same COLOUR as the wire.
Good idea about the NO power bars.
Is it possible to physically keep the 2 voltages separate, ie 110 from below the floor and 230 from above in the ceiling?
I think the Classics are on the verge of inter-unit communication, ie master/ multiple slaves.
Maybe the use of 12v, plus 24 or 48V systems matched to the needs, allows for later additions of B. bank if needed but keeps cost to minimum now using a smaller bank. Slightly more complex but designed for specific portions of demand.

hth

cruiser guy

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

I currently have an unused 600 watt modified sine wave inverter than could be used to power lights. It's cheap and its here. This is a 12 volt 230vAC 50 Hz unit so it would work with the local power here. The 110vAC is where the medical stuff will be initially so it should have something more like a true sine wave inverter. I understand the modified sine wave inverter will make lights hum and can make them harder to start but that's the price to pay for a cheaper inverter.

I'll look at working out a system that is minimal at this time with the intention that most of it will be redundant when we do expand at some point in the future. I'll keep wiring from the inverter to the panels to a size that can reasonably be expected to carry whatever the final load might be.

The Grundfos SQFlex operates on anything from 30 to 300vDC and 90 to 240vAC according to the literature I have here from Grundfos. I have read in some of the older posts in the water pumping section about the SQFlex not working well on voltages in the lower section of the allowable range.

I'll push Grundfos for the wiring diagram for incorporating a charge controller and then look at pricing a charge controller that can take high solar array voltages and with multiple battery bank voltage settings. I'll also look for one that can communicate with other charge controllers.

My concern with a minimalist system is that it will quickly be overwhelmed and we'll start burning out inverters and the like. It also requires a transfer switch to go from battery to generator while the other more expensive inverters can handle this switch automatically. Not a deal breaker in any of that, just the drawbacks that I see.

Cariboocoot

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

An overloaded inverter won't burn out, it will shut down. Possibly over and over and over. Usually they'll shut down before any circuit protection trips.

Getting the small system for lighting up and running is probably the best first stage you can go with. There's two issues: the unknown loading resulting in incalculable run time and the Voltage difference between array and inverter. That last one presents quite a problem, especially if the array is running 210+ Volts. That's MidNite Classic Lite 250 territory: a $600 charge controller. Or worse, the Xantrex 600 Volt controller for double the money. At those prices it may be cheaper and certainly easier to buy dedicated panels/controller for the 12 Volt system instead.

I'd make it so that this small inverter is all hardwired to lighting so that nothing can be plugged in to it, just in case. Then when you know what else needs to be powered you can install an appropriate higher capacity system for those loads (with dedicated outlets).

niel

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

actually an inverter could burn out under heavy load if there is any problems with the fets themselves or possibly other parts.

cruiser guy

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

If we go with the fancy high power charge controller can it still be chained with a compatible controller running at a lower array voltage at a later date? After taking more with the medical staff here I think I'm going to go with a 2000 watt 110vAc 60Hz pure sine wave inverter as our minimum as they are talking single items of equipment with a 500 watt load. Unfortunately that is a nearly $2000 item that will be redundant with a 48 volt battery bank in the future.

It's not just lights that we're looking for but also lab equipment as there is no medical lab of any kind in the area either. Lights will be just a couple.

Additional power needs, such as an emergency at night will require generator operation with this lower cost system.

westbranch

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

a bit of sleuthing may be in order with that medical equipment, you will need to investigate if there is any inrush or surge when the item is turned on. motors for example can be 3 x or more the rated consumption... you might find one of those larger more costly inverters is the best deal all round.

hth

Cariboocoot

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

niel wrote: »

actually an inverter could burn out under heavy load if there is any problems with the fets themselves or possibly other parts.

All the inverters I've seen will shut down automatically if overloaded. That is not to say there aren't any out there that would fry instead.
Running them at or near maximum capacity however is bound to shorten lifespan.

Cariboocoot

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

cruiser guy wrote: »

If we go with the fancy high power charge controller can it still be chained with a compatible controller running at a lower array voltage at a later date? After taking more with the medical staff here I think I'm going to go with a 2000 watt 110vAc 60Hz pure sine wave inverter as our minimum as they are talking single items of equipment with a 500 watt load. Unfortunately that is a nearly $2000 item that will be redundant with a 48 volt battery bank in the future.

It's not just lights that we're looking for but also lab equipment as there is no medical lab of any kind in the area either. Lights will be just a couple.

Additional power needs, such as an emergency at night will require generator operation with this lower cost system.

If you're going to shell out $2,000 for an inverter and want a 48 Volt system anyway: http://www.solar-electric.com/fx3048t.html

Kind of like jumping in with both feet, but if the money is the same there's no point going for a smaller inverter to begin with if you know you'll need more power later anyway.

You can connect more than one charge controller to the same battery system, regardless of array Voltage, so long as the output Voltage can be set the same.

cruiser guy

Re: Power ideas for off grid hospital in Sierra Leone, West Africa.

Cariboocoot wrote: »

If you're going to shell out $2,000 for an inverter and want a 48 Volt system anyway: http://www.solar-electric.com/fx3048t.html

Kind of like jumping in with both feet, but if the money is the same there's no point going for a smaller inverter to begin with if you know you'll need more power later anyway.

That's my thought as well but I'm not the one holding the purse strings. All I can do is present the case as I see it and let those who decide how to spend the money decide what will ultimately be done. When they decide then we move forward and make it happen. Later when they ask why we are decommissioning a $2000 inverter I bring them back to this discussion!

Anyone have a source for a low cost 48v to 230vAC 50Hz inverter, modified or pure sine wave of about 600 to 1000 watts? I know about the Outback Power VFX series that starts at 3kW but that's right back to my original system that can be built upon.