4.10 Installing the charger previous
4.10.1 Introduction index page

The inlet, canal and charger site are very important points if the charger will be installed there permanently. However, when it is just for testing or a demonstration of a few hours to a few months, a lot of short cuts are possible:

In the subsequent paragraphs, the inlet, canal and charger site are dealt with as if the charger will be installed permanently. This has been done because such permanent solutions will be needed in case the firefly system will be introduced for real after a demonstration project. Please bear in mind that cheaper and less time-consuming options can be used for demonstration purposes.

With respect to the inlet and the canal, it makes a lot of difference whether there are irrigation canals in the area and local people clearly have the skills to plan, build and maintain them. If irrigation canal technology is known in the area, local people probably know much better how to plan, build and maintain an inlet and canal that will function in their environment. Then there is no need for an outsider to tell how or where to build an inlet or canal, so you could skip the corresponding paragraphs. However, don't expect that the charger can be connected to any suitable irrigation canal because farmers will claim the water in such a canal, see next paragraph.

If irrigation canal technology is not known in the area, the firefly project will have to bring this in as well, see par. 4.10.3 and par. 4.10.4. In these paragraphs, only major points are discussed. It would go too far to discuss in detail all subjects related to estimating whether a creek has enough water in dry season and planning the inlet, canal and site. Literature is available on these subjects, e.g. HARVEY, 1993. Do not underestimate the importance and complexity of these issues since they could easily lead to a disappointing performance of the whole project. A charger connected to a dry creek, with a washed-out inlet structure or a silted canal will just not work.

The first problem that arises when there is no irrigation technology around, is choosing the creek or river that will be used as a source. Of course local people should be asked, but their memory might prove unreliable once it comes to estimating what is the lowest level in dry season, whether a creek will fall dry every dry season or only in exeptionally dry years etc. Local people might also give the answers you want to hear so badly just to please you or because they want the project to push through. Generally, people who use the water in a creek, will know more about it, so maybe ask the women who use it for washing, or the children who go there for a swim.

When information of local people is not conclusive, you'll have to find out yourself. There are several handy methods of measuring the flow of a creek at a one moment (see e.g. HARVEY, 1993) but the problem is to estimate how the flow will vary over the year. That is why estimation methods based on rainfall data, soil type etc. could prove even more useful. Also there are handy methods to process flow data into power output for a certain type of micro hydro plant, and evaluate what capacity plant would be most advantageous. In general terms, this issue was discussed already in par. 3.1. Not only the low flow figures during dry season should be known, but also the highest water level that is reached during floods. The strength of floods determine what type of inlet is most appropriate, what is a safe site for the charger etc.


4.10.2 Linking in with irrigation technology

If there are irrigation canals in the area, maybe the charger can be connected to an existing irrigation canal and then the inlet and canal are virtually ready for use. However, problems are to be expected with the farmers using this canal. A logical solution would be that the firefly project guarantees that the capacity of the first stretch of canal will be increased so much that the farmers further along the canal will always get as much water as before or even more (of course this only makes sense when it is the capacity of the canal that is limiting the water supply, and not a too low dry season flow in the creek it gets its water from). Then the maintenance work of the first stretch should be shared equally by all water users.

However, farmers could demand an exorbitantly high compensation or just want this thing off their canal. Take the following issues into account:

If it is about a permanent site for the charger, preferably the firefly user group should negociate directly with the affected farmers. Having a representative from a development project around at such a meeting could be counterproductive because then the easiest way out for all concerned is to ask more money from the development project, irrespective of whether the demands of the farmers are fair. Such a request would probably cause delay (it has to be decided upon and if granted, the money has to be found) and if it is refused, the negociations have to start all over again. When the request would be granted, it sets a precedent and with subsequent problems, users might come again to the development project for a solution. Also other user groups that want to start later on, might expect a similar amount of money from the development project.

The issues mentioned above could make it practically impossible to use an existing irrigation canal as a source. Then options are:

When discussing a firefly scheme with local people, it might be difficult to explain what it is about. Some people might think that the charger just needs running water and that it can be thrown in the river to have it work. They will not object to this since water down in a river is useless for them. But water in an irrigation canal is a totally different matter and they might not like it at all when foreigners come with ideas that interfere with this water.

Local people know all about height differences and amounts of water, but not in terms of meters of head and liters per second of flow. So to explain what the charger needs, more familiar terms have to be used:


4.10.3 Designing an inlet

The inlet site, the course of the canal and the site of the charger are mutually linked. One could start with an advantageous inlet site, plot out the course of the canal with the desired slope and where the canal is high enough above the flood level of the creek or river, the penstock starts and the charger can be placed down there. Or one could choose a good site for the charger, plot out the penstock and the canal, and the inlet will be there where the canal reaches the creek. In reality, one could work out several different alternative plans (inlet, canal and charger site) and choose the most advantageous one.

Important aspects for choosing the inlet site are:

Generally, an inlet consists of:

There is also the option of using temporary weirs that are likely to be washed out by floods in wet season, but that are cheap to rebuild using locally available materials and labour.


4.10.4 Designing the canal

Important aspects for the canal are:

Designing a canal involves hydraulic calculations that are too specialistic to discuss here, see HARVEY, 1993. Only a simple comparison will be given.

An important feature of the canal is its slope. In irrigation schemes or Micro Hydro schemes, generally the slope is chosen such that the water velocity in the canal will be a compromise between two evils:

Because the firefly charger needs such a low flow, the canal will have a small cross-sectional area and even for a velocity of only 0.3 m/s, the slope will have to be some 3 %. This would be acceptable only if the slope of the main river is high as well so that the head loss in the canal is still low compared to the head at the charger site.

In cases where the slope of the creek or river itself is only a few %, this is not practical. Too much of the available head will be lost in the canal, therefor the canal must be made longer for obtaining the head the charger needs, this in turn makes tha seepage losses increase considerably etc. By reducing the velocity to 0.2 m/s, a slope of only 1 % is needed. This means that the canal will get silted and needs digging out regularly. Since the flow is low, also the amount of silt carried in by it will be low and this should not be such a problem. Anyway it saves the costs of building a large forebay tank that serves as a silt basin and emptying that regularly.

Another consequence of choosing a low velocity is that the canal needs to be wider and deeper, see table below:

Velocity: 0.3 m/s 0.2 m/s
Flow (8 l/s for charger, + 5 l/s for seepage losses and safety factor): 0.013 l/s 0.013 l/s
Bottom width: 0.19 m 0.23 m
Height (this is the height of the water, the top of the wall should be at least 0.15 m higher) 0.23 m 0.28 m
Top width (canal walls have slope of 100 %): 0.65 m 0.79 m
Roughness of canal with short vegetation, corrected for canal height: 0.084 0.076
Slope 3.3 % 0.9 %

When a canal has to be designed and plotted out in the field, it is worthwhile to get proper literature (for instance HARVEY, 1993) and make such calculations yourself. This would also make it possible to compare alternatives properly. Or ask advice from someone who is familiar with designing canals, e.g. an engineer from the dept. of irrigation, a civil engineer. Also it could be worthwhile to visit an area where irrigation canals are functioning. Topographical maps (if available), landscape pictures and details from inlet sites, charger sites, the terrain the canal has to cross etc., are helpful for documenting and for explaining to other people about the plans.

The following could serve as a rough guideline:


4.10.5 A site for the charger

Important aspects for choosing the charger site are:


4.10.6 Linking the parts together

The two main things a charger needs are a head H and a flow Q. There are minimum requirements for these (see par. 4.1), but once the head is above 7 m, the charger becomes quite flexible. It can generate the same electrical power with a relatively high head and low flow, or with a lower head and higher flow. In this paragraph, some general guidelines will be derived regarding the planning of inlet, canal and charger site. In this, many of the site-specific issues that were mentioned in the previous paragraphs, are ignored. Also guidelines that are valid in all cases (e.g. high slope of river Sr, low flood allowance F, low seepage losses constant Cs are favourable in all situations) are not mentioned again.

To work things out, the normal water level at the canal inlet can be used as a reference height. Clearly the head H the charger experiences, is the difference between height hs of the charger site itself and height hf of the forebay tank in the canal just above the charger (the forebay is the place where the water enters the penstock pipe, see par. 4.10.7):

H = hf - hs

From the canal inlet (the reference height, so height = 0), both the river and the canal go downhill so to keep things logic, for both the slope of the canal Sc and the slope of the river Sr, negative figures are used. Also the height of the forebay hf and the height of the charger site hs are negative compared to the reference height of the inlet.

The slope Sc of the canal is much smaller (less negative) than the slope Sr of the river and that is why downstream of the inlet site, the height difference between the canal and the river becomes bigger and bigger. In practice, the charger must be protected against floods by placing it well above the normal water level of the river at that place: The flood level allowance F.


hs = Sr * Lr + F


hf = Sc * Lc


H = hf - hs = Sc * Lc - Sr * Lr - F


hf = Height of forebay (negative)
hs = Height of charger site (negative)
Sc = Average slope of the canal (negative)
Sr = Average slope of the creek or river (negative)
Lc = Length of the canal
Lr = Length of river between inlet site and charger site
F = The `flood level' allowance, the height of the charger above normal water level in the river at that site

Along the canal, part of the flow Qi that entered the canal, is lost as seepage losses Qs. Roughly speaking, these seepage losses Qs will be proportional to the length of the canal Lc. So:

Q = Qi - Qs = Qi - Cs * Lc


Q = Flow at charger
Qi = Flow at inlet
Qs = Flow lost as seepage losses
Cs = Constant representing seepage loss per m of canal length, Cs depends on soil type and wetted part of canal cross-section, which in turn depends on the slope Sc of the canal and the flow Ql at that point.

In reality, seepage loss per m canal Cs is not constant, but depends a.o on the flow Ql at that point in the canal, which in turn depends on the seepage losses from the inlet up to that point. By taking a constant value for Cs, this effect is neglected.

Using these formulas, some general guidelines can be derived for planning the inlet, canal and charger site, see table 4.3. Depending on what criteria one sees as most important, different guidelines will come out.

Table 4.3: Guidelines for choosing the inlet and charger site, and designing the canal.
Criterium: Guidelines:
Minimum canal costs Canal costs will depend mainly on canal length Lc and to a lesser extend on the cross-section of the canal. Canal length Lc will be minimal if:
  • The head H for the charger is low. The minimum is 5 m but then the charger will produce only 100 W electrical power. For 165 W, a head of 7 m is needed, see par. 4.1.
  • The slope of the canal Sc is low.

If one would like to keep the cross-section of the canal minimal, completely opposite guidelines come out:

  • The slope of the canal is high (for low canal length, there is the opposite demand so an optimum must be found).
  • The flow Q of the charger will be low. However, a low flow means that head must be high to obtain the desired electrical power.

Generally, canal costs will be lowest if canal length is low, so if head is minimal and flow is relatively high.

Minimum penstock costs Penstock costs depend on diameter and length of pipe that is needed. When head is high, a small diameter will do but a long length is needed, while for a low head, a short length of a larger diameter pipe is needed. Generally, pipe costs will be about the same in both cases. But in concrete cases, quite a lot of time and money can be saved if:
  • The price of one diameter pipe is low compared to other diameters (compare the price per l/s of flow it can conduct, see par. 4.10.8).
  • By choosing a slightly lower flow (so higher head), a smaller diameter pipe can be used.
  • A certain diameter pipe is sold per m or in lengths of 6 m, while another diameter pipe is sold only in complete rolls of 60 m (so: much more than is needed for one charger).
  • A local shop has one diameter in stock while another diameter has a delivery time or has to be bought in a major city and transported.
Minimum maintenance costs Maintenance costs for the canal depend strongly on the length of canal. It should be weeded regularly and when there are leakages or small landslides, they should be repaired. From this point of view, a short canal (so: low head and high flow) is preferable. Another maintenance job is digging out silt that has been deposited in the canal. This is only significant if, at times, the creek or river carries a lot of silt. To keep this to a minimum:
  • The slope of the canal could be chosen so high that the velocity is above 0.3 m/s and silt is not deposited in the canal (instead, it will deposit in the forebay / silt basin, but from there, it can be flushed out or shoveled out more easily). For a firefly system, this is not a practical solution because then such a high slope is needed that the canal has to be made significantly longer, see par. 4.10.4.
  • The amount of silt that enters the canal is proportional to the flow at the inlet Qi. By choosing a long canal, the flow can be less and, with that, also the total amount of silt to be dug out. Only the first stretch of canal will become silted. By making this part deeper and wider, it needs digging out less often and the digging can be done more efficiently.

Generally, maintenance will depend strongly on local conditions. The best way to reduce maintenance requirements is by choosing a smart inlet site, easy route for a canal and constructing everything carefully. To make that maintenance is easier and is needed less frequently, the canal could be designed for a higher flow, so that it will still function satisfactorily when it is partially overgrown or silted.

Minimum water use Reasons for designing the system like this could be:
  • Low dry season flow in the creek.
  • The charger uses water from an irrigation canal.

For producing the desired electrical power with a low flow Q, the head H must be high. To obtain such a high head, the canal must be long.
If seepage losses are very high, it makes no sense to make the canal any longer than a certain length. Then the charger itself would need a lower flow Q because of an increased head H, but the flow at the inlet Qi would hardly go down since seepage losses Qs increase with canal length.
If the same canal is used for irrigation, it is always best to choose for the highest head since the seepage losses would be there anyway.

Possibilities for later expansion This could be desirable if more users want to join in later.
For generating more power, either the head H or the flow Q has to be enlarged later on:
  • The head H could be increased later on by making the canal longer, which is only practical if agreement can be reached with the land owner, there are no difficult stretches to cross etc. It also means that the charger must be moved downstream so there should be a suitable site there.
    Of course the head could also be increased by building a new inlet upstream of the first one and digging a canal from there, while the charger remains at the same site. This is impractical because it involves too much work.
  • The flow can be increased easily by making the canal wider and deeper. Since the slope of the canal will remain the same, the velocity in it will increase so it needs to be only a little bit deeper and wider to carry a much larger flow.

Of course also the capacity of the charger should be increased:

  • When the head H is quite high (say 10 m or above), the capacity of a standard charger can be increased substantially by just fitting a smaller blocking timber or removing it altogether. Then flow will increase and a larger diameter penstock pipe might be needed. Also the switchboard should be modified so that 2 or more batteries can be charged in one go. In case there is only one battery to be charged, the flow to the charger should be limited, for instance by partially opening the valve in the forebay tank, see fig. 4.32. See also annex B: *Design for a large capacity charger.
  • When the head is small but the flow could be increased further, either 2 or more standard chargers must be installed, or a special low-head charger must be used, see annex B: *Design for a low-head charger. This is a less attractive option.

Generally, expansion possibilities are excellent if:

  • The head is already high or if the canal can be made longer easily.
  • The flow can be increased further. This is the case if even in dry season there is plenty of water in the creek or river, and if there are no other uses (irrigation) for this water.
Some general comments:
  • If a canal has to be constructed for the charger, canal costs will be a major issue and the design of the scheme should be such that canal costs are minimal. So the canal should be short and consequently head will be low.
  • If an existing irrigation canal will be used, the flow consumed by the charger will be most important. Then generally a charger site with a high head (so that flow can be low) will be most advantageous.
  • In practice, there won't be that much to choose about the inlet and charger site and the course of the canal. It will be more about avoiding problems and obstacles, e.g.:
    • a creek that might fall dry
    • farmers claiming all the water from a certain canal or creek
    • land owners who don't allow a canal to be dug through their land,
    • very steep, rocky stretches where it would be hard to chisel out a canal, or on the other hand: Sandy stretches where seepage losses would become excessive and erosion risk is high.
    • Distance to homes of users, distance to someone who might keep a look on the charger.


The standard charger is quite flexible in different ways:

This flexibility can be used to make the most out of the opportunities that are offered by the natural environment in the area. Also with regard to organisational aspects, this flexibility can be used. A small group of users can start, charger capacity could be increased if more people join in, eventually a number of users living a bit far from the charger might split off and start with their own charger etc. This all could be done without writing off the investment in the original charger and the penstock.


4.10.7 The forebay tank and screen

First, an expensive, permanent design will be discussed. See the end of this paragraph for a cheap, makeshift solution.

Before the water enters the penstock pipe, it should pass a screen with about 5 mm mesh (of galvanised iron wire, or stiff plastic). This is to protect the charger against pieces of wood or any other foreign object from entering the pipe and getting stuck between the nozzle and the runner. Without a screen, playing children might even let pebbles be sucked into the pipe. The larger the surface area of the screen, the longer it can work before it gets blocked by water plants, dead leaves etc and needs cleaning. Of course this also depends on how much water plants etc. are carried along by the water. A screen area of 0.1 m per liter/second of flow worked well in the Cambulo project, but it still needed cleaning every day. So for the standard charger, a screen area of about 0.8 m should do.

The shape of the screen does not matter much as long as it is easy accessible for cleaning. It could be fitted directly onto the penstock pipe. Then it could be like a long cylinder, or have a square or rectangular cross section, see fig. 4.32. It could also be a flat screen fitted to a wall of the forebay tank, with the penstock starting from the cavity behind the screen.

Fig. 4.32: An expensive, permanent forebay tank.

If the screen needs cleaning very often, it could make sense to have a removable screen in front of it. With a screen like in fig. 4.32, one could tie a bag of woven plastic (like the ones for rice etc.) around it, take it off once it gets clogged and shake off all dirt.

Apart from the screen, the forebay construction of fig. 4.23 has the following features:

In fig. 4.32, it looks like the penstock pipe leaves the forebay tank above the canal for excess water. In reality, it should be fitted besides the canal for excess water, so that it can bend downwards to the charger.

The top end of the penstock pipe should be inclined 5 to 15 downwards. If it would be completely horizontal or even inclined upwards, a large bubble of air gets trapped in the penstock pipe. Then the charger might still get quite some water, but not enough head (= water pressure) since there is not enough weight of water in the pipe to produce this head. With the top end slightly inclined downwards, the air can come out at the forebay end.

A pipe going straight down from the forebay tank to the turbine is no good either since this might cause a `water hammer' effect. Once the valve is closed, the pipe might be filled so fast that a mere `wall of water' starts falling down the pipe. Once the water reaches the nozzle, it cannot escape that easily and the whole column of water is forced to slow down suddenly. As a result, there is a pressure surge that makes the charger and pipe jerk. It could damage the pipe, the turbine or knock the charger off its foundations. This is why the top end should not be inclined too much, 15 seems a safe value.

A penstock pipe filled with water is quite heavy so it should be fixed properly:

The length of the rear leg of the charger can be adjusted, so that it is well-aligned with how the pipe comes in. If the turbine is not stable enough when it is just standing on its legs, fix the legs on boards or logs (but make sure that the charger can easily be moved to a safe place when the place where it stands could be flooded). If the soil is soft, this might also be necessary to prevent that soil is washed away by the water from the turbine or that the whole site changing into a mudhole.

If the charger will be installed only for demonstrating for a few hours to a few months, there is no need to build an expensive, permanent, concrete forebay tank. The only thing that is needed is a cylindrical screen that fits directly on the penstock pipe. For connecting the charger to a small, existing canal, choose a stretch where the downhill canal wall is wide and strong. Place the pipe with screen over the canal wall with its opening upstream. Maybe dig out a bit from the wall where the pipe crosses it, so that the pipe will be a little lower there (take care to repair this properly when the charger is removed later, it might start to leak and erode away the soil underneath). Tie the pipe to a tree or a pole, so that it won't slide down once it is filled with water. Then build up the canal wall upstream from where the pipe crosses it with soil, sods, stones etc. The wall should be so high that the water can be raised until the screen is completely submerged. Finally build a dam in the canal right behind the screen so that the water level upstream of it rises, the screen is submerged and the charger starts to run. This dam will serve as an overflow for excess water so it should be made of stones or sods and not of mere soil. If conditions at the site are favourable and there are some people to help, the charger can be installed in half an hour.

If a canal is dug especially for the charger, a similar simple forebay structure can be made of soil, sods etc. Then take care that excess water does not erode away stable soil that should serve later as a foundation to the canal should continue. It might stimulate people to see the charger run already before the canal is finished. Once the head is 3 m or higher, the charger should deliver a little bit of electricity.


4.10.8 The penstock pipe

As a penstock pipe, PE (Poly-Ethylene, a kind of plastic) pipe can be used. This is the black plastic pipe that is often used for potable water systems. For one outer diameter, pipes with a different pressure rating could be sold. These have a different wall thickness and consequently also a different inner diameter. Normally, even the pipe with the thinnest wall thickness should do since pressures in a potable water system usually are higher than in a firefly system.

PE pipes are light and slightly flexible: In the Philippines, diameters up to 2.5" were sold on a large roll like a hose. When it is bent too much, it produces a sharp hitch where the pipe is almost flattened. Such a hitch is difficult to get straightened out again and it might be necessary to cut it out. PE pipes can be joined by heating the contact surfaces and then holding them together, but proper tools and experience are needed for this. Connections can also be made by making a piece of steel pipe that fits inside the PE pipe and then securing the PE pipe with iron wire or a hose clamp.

What diameter pipe is needed, depends on the flow the charger will consume:

Pipe diameter rating (in inches) Maximum flow (in liters/second) if inside is:
smooth moderate rough
2 " 3.6 3.0 2.6
2.5 " 6.6 5.5 4.7
3 " 10.7 8.9 7.6
3.5" 16.0 13.4 11.5
4 " 22.9 19.1 16.3

One would expect that plastic pipe would be as smooth at the inside as at the outside. However, in the Philippines, PE pipes were sold that were definitely not smooth at the inside. That is why also values for less smooth pipes are included.

The calculation of the maximum flow ratings are based on the following:

Length of pipe: multiply maximum flow by:
1.5 times head 1.2
2 times head 1
3 times head 0.80
4 times head 0.72

If these maximum flow ratings are compared with the flow the charger will consume at different heads (see par. 4.1) it follows that, roughly speaking, the following pipe diameters will be needed for different heads:

Head: Pipe diameter:
5 - 10 m 3 "
10 - 18 m 2.5 "
more than 18 m 2 "

These values are valid in case such a blocking timber is fitted that power output will be 165 W. If one chooses a smaller blocking timber so that the charger will produce more power, the flow will be higher and also a larger diameter pipe is needed.

The above guideline is rather rough. It is better to inspect the site, measure the head and the length of pipe that is needed, check whether the inside of the locally available PE pipe is smooth, and make your own calculations.

It is not adviseable to go above the maximum flow ratings that are calculated this way since then friction losses increase very fast and power output could become disappointingly low. In case the actual flow is coming close to the maximum flow for a certain pipe diameter, length and roughness or even surpasses it a little, consider the following options:

The maximum allowable flow values for a pipe with a certain diameter, roughness and length ratio were calculated based on the assumption that friction losses are 20 % of total power. Since friction losses increase with the square of the flow, it is possible to estimate friction losses by comparing the actual flow with the maximum flow for a certain pipe:

Pf = (Qa / Qm) * 20 %


Pf = friction losses, as a % of expected power without friction
Qa = actual flow through pipe
Qm = maximum flow for such a pipe

Friction losses were not included in the calculation of expected power output in par. 4.1.

Pipes are classified by their diameter in inches but this inch rating gives little information on the actual dimensions of the pipe. Even the inner diameter can be 3 % larger than the rated diameter in inches. Pipes for high pressure could have the same outside diameter than low pressure pipes, but a greater wall thickness. Consequently different types of pipes with the same inch rating could have slightly different inner diameters. So for making the steel connection pipes, you will have to measure the inner diameter yourself once you have found a suitable pipe. Often PE pipe is not completely round so measure the largest and smallest diameter and take the average of the two as the real diameter. See par. 4.6 for how to calculate and build steel pipes that fit tightly into plastic penstock pipe.

Connection pipes are needed for:

The penstock pipe forms a major part of the charger costs so it is worthwhile to try to economise on it. When local hardware shops don't have the size in stock that is needed, or don't have any large diameter PE pipe at all, also delivery time becomes an issue. And finally, it might be impossible to buy just the length you need since hardware shops could insist on selling complete lengths and complete rolls of pipe or ask exorbitant high prices if you want to buy per meter. It might be possible to economise on penstock costs by:

There are also other kinds of pipes that can be used for the penstock:

index page