Small Hydro power Siting
Paul Cunningham

There are small streams occurring over much of the countryside.  Perhaps you
are wondering if a brook in your area is suitable for developing into a power
source.  The following is intended to show the procedure I used in my own case
to arrive at solutions to various problems.  Hopefully, discussing the thinking
involved will provide some insights.
A small scale water power system requires a more specific site than either a
wind or photovoltaic one.  You do need to have some flowing water.  On the
other hand, it isn't necessary to have very much, or much pressure, and it
doesn't have to be very close to the point of use.  My situation will
illustrate this.
Here in the Canadian Maritimes it is difficult to go very far without finding
some type of stream.  I live in an area of rugged topography which enhances the
water power potential.  My house is located near a brook that most times of the
year has a fairly low flow rate.  There is normally little water in the stream
above the house while water from springs which come to the surface steadily
increase the flow as the water runs downhill.
One logical place for the intake and beginning of the pipeline is near my
house.  Although flow increases further downstream, the slope decreases.  Near
the house the brook drops around 8 feet for every 100 horizontal feet.  So
running a pipeline downstream 1,000 feet produces a combined drop or "head" of
75 feet.  This looked like a reasonable place to start although the site
permits running a pipeline 3,000 feet before the brook meets another one
running almost level.
Polyethylene pipe was purchased (in 1978) and simply laid on the ground.  A
small screened box serves as the intake and was set in the brook with a "dam"
of earth and rocks sufficient to raise the water level about one foot.  At this
site, the maximum power will be produced at a flow rate of about 20 gallons per
minute (GPM).  This is the point where the dynamic (running or net) head is
equal to two thirds of the static head.  So there will be 50 feet of net head
at the end of the pipe when the water is running with a suitable nozzle at the
end.
Any increase in flow will result in a decrease in power available due to
increased pipe friction losses.  Right away then one third of the precious
power potential is lost.  At lower flow rates the pipe loss decreases which
results in an increase in efficiency as flow decreases.
So why don't I use a larger pipe?  Well, it costs more and sometimes 20 GPM is
all there is in the brook.  Also a larger pipe would aggravate the problem of
freezing at low temperatures with no insulating snow cover.  This is because
the residence time would increase with larger pipe.   The water entering the
pipe is (slightly) above freezing, in my case, and cools as it travels along
(when temperatures are very low).
So why don't I  bury it?  Yes that would be nice and hopefully I will when I
can afford that and larger pipe too.  It is a case of the shoemaker being
inadequately shod as I content myself with the present system.  Besides, it has
spurred me on to other possibilities that we will look at later in future
articles.
Back to the 20 GPM at 50 foot head.  A 3/8 inch diameter nozzle is about the
right size for this, giving 19 GPM  According to the spouting formula the
velocity of a jet of water will be:
V = 2gH = 2*32.2*50 = 56.7 ft./sec.

g = 32.2 feet/sec/sec (acceleration of gravity)
H = head expressed in feet

How much potential power is this?  A U.S. gallon of water weighs 8.34 lbs. and
the flow is 19 GPM; then  8.34 lbs. per gallon X 19 gallons per minute = 158
lbs. per minute.  Now, 158 pounds of water per minute falling 50 feet has 7,900
foot-pounds/minute of energy (simply multiply the factors).  Conversion to
horsepower is accomplished by division by 33,000., thus 7900/33,000 = .24
horsepower.  Since 746 Watts of energy is equivalent to one horsepower, .24 hp.
X 746 Watts per hp. = 179 Watts of potential squirting out the nozzle.  This
means that the potential power was .36 horsepower or 269 Watts before going
through the pipe.  Since nozzles tend to be very efficient not much loss is
expected.  But keep in mind that every time the energy goes through a change,
power is lost.  All right, how about a 9 Watt loss to make an even 170 Watts.
This may appear a little sloppy.  But you must realize that these systems do
not have to be very precise-- they are quite forgiving.  Also many of the
measurements are difficult to determine with high accuracy.  So close
approximations are sufficient.
Thus far things are reasonably straightforward - a pipeline with a nozzle at
the end.  Now what?  Conventional practice would suggest some sort of impulse
turbine such as a Pelton or Turgo.  It would also be possible to use a reaction
machine.  It would have to resemble one of those spinning lawn sprinklers
rather than say, a propeller type.  This is because of the very small nozzle
area.  The impulse type looked easier to build.
At this site it is necessary to send the power back upstream 1,000 feet to the
house.  I wanted to use 12 VDC and wanted some way to transmit the power other
than the very large wire that would be required at this voltage.
In the spring, when the flow in the brook was very high, various 12 VDC
generators were operated with the pipeline ending near the house.  But this
could only be temporary, as ways of solving the transmission problem had to be
discovered.  Of course using wires wasn't the only possibility.  I could always
charge batteries downstream at the generator and then carry them up to the
house.  Or perhaps a reciprocating rod kept in tension could be used to
transmitted the power.  But all things considered, producing electricity at a
voltage higher than 12 VDC looked the easiest.
I thought generating AC electricity at 60 Hz. like regular commercial power
would permit using standard transformers and make it easy to change the
voltage.  For this I bought a "Virden Permabilt" 120 VAC generator.  This
produces 120 Watts rated out put and 60 Hz. at 3600 RPM.  These machines are
reworked DC auto generators with rewound field and rotor with a slip ring and
brush to carry the output.
An impulse turbine should have a surface speed of about half the jet velocity.
So at the 56 feet per second a turbine wheel slightly less than 2 inches in
pitch (hydraulic) diameter is required.  This is a little on the small side but
I did make a Turgo wheel of this size so the rotational speed would be right
for direct drive.  Yes it's possible to use speed increasers with a larger
turbine but I didn't think there was anything to gain and only power to be
lost.  It turned out that the alternator would not generate 120 VAC at a low
power level.  The field required 10% of the rated 1200 Watts output to put out
120 VAC regardless of the load.  Therefore a lower output voltage was necessary
to properly balance the system.  It was determined that under the site
conditions an output of 24-25 Watts was required to be in the correct ratio:
120 VAC/10 Amperes = 24 VAC/2 Amperes or 48 Watts.
Now you are probably wondering how come only 48 Watts is being produced.  Well
that is what that combination of turbine and generator put out.  And this isn't
the end either.  Next the juice went through a 25-110 volt transformer, through
1000 feet o f 18 gauge wire (two strands), another transformer down to 12 volts
and then through rectifiers to give DC.  In the end only 25 Watts or about 2
Amperes actually found its way to the battery.
This setup didn't last long enough to make many improvements.  It was hard just
keeping it alive.  The alternator used only one slip ring.  The other conductor
was the bronze tail bearing!  Both items had limited life under 24 hour
service.  Besides the efficiency was low anyway.
INSERT HYDRO ILLUSTRATION
I still needed a reasonable system.  At least one with a longer life.  In the
next attempt a 4 inch pitch Pelton Turbine was cast in epoxy using a silicone
rubber mold.  This directly drove a car alternator with a rheostat in series
with the field to adjust the output.  Transformers (3) were connected to the
three phase output to raise the voltage for transmission with the (now) 3-18
gauge lines.  Then a similar set of three transformers is used at the house to
lower the voltage and a rectifier to make the DC conversion.  About 50 Watts
was still generated (4 Amperes 12 volts) but more made it into the battery--
about 3 Amperes.  The reason for this is the automotive alternators have more
poles (12 Ford, 14 Delco) and generate at a higher frequency.  This improves
the efficiency of small transformers even though they are "designed" to work at
60 Hz.  Now the system has an efficiency of around 21% (36 Watts/170 Watts)
using the power available at the nozzle as the starting point.
Three Amperes in a 12 VDC system doesn't sound like much.  But this is
sufficient to run the lights, a small fridge (Koolatron) and a tape player-
radio.  My house is small and so are my needs.  There was sometimes even extra
power and I could run Christmas lights or leave on things just to use the extra
power.
At some point it occurred to me that I might generate more than electricity if
I could produce turbines for others in a similar situation.  Peltons were made
first for sale.  Originally these were made of epoxy and later of a high-
strength and abrasion resistant Polyurethane.  This endeavor  busied me some
but it soon became apparent that to survive doing this sort of thing would mean
producing complete generating units.
Turgo turbines looked more reasonable than the Peltons for this, due to their
greater flow handling capability for a given size.  Using a 4 inch pitch
diameter turbine wheel allowed as many as four one inch diameter nozzles to be
used.  This resulted in a very versatile machine.
The first production models used automotive alternators (Delco) since they are
inexpensive, dependable, available and most people wanted 12 VDC output.  But
these couldn't operate with heads of less than 20 feet or so.  Also the
efficiency of these alternators is in the 40-50% range and I thought there was
room for improvement.
Back in the R and D department, work was proceeding to develop a better
machine.  The Turgo turbines operate in the 60-70% efficiency range.  These are
made in re-usable silicone rubber molds.  This placed certain constraints on
their design and so limited the efficiency.  But other tests showed there
wasn't much to be gained by changing the shape of such a small wheel.
However, the generators used so far had efficiencies in the 50% range or less.
They also had electric field coils which made for easy adjustment of the output
but also took part of the output to operate.  It looked like the use of a
permanent-magnet (PM) field would be a help and could make operation at very
low-heads feasible.  Yes, DC motors with PM fields could be used as generators.
But my experience with machines where brushes carried the full output was
disappointing.  Longevity was a problem -- remember these are going to run 24
hours a day.  If alternating current could be generated then transformers can
be used to alter the voltage to suit the site.
It is well established that the most efficient generator type, especially in
small sizes and at low speeds, is the PM-rotor alternator.  Just like a bicycle
generator.  There is also nothing to wear out besides two ball bearings.  That
would be a feature and a half.
After a few tries, standard induction motors were used by keeping the stators
and building new PM rotors.  This produced a machine capable of generating
power with an efficiency of over 80%.  Standard 60 Hz. AC output was possible
at 1800 RPM for these 4 pole machines.  Experience suggested that frequencies
of 50-400 Hz. would operate standard transformers quite well.  This, combined
with the recommended output wiring, produced a machine able to generate almost
any voltage.
So how is it looking back at my site?  Using the new PM rotor alternator about
100 Watts of power is produced.  This is an efficiency of 100 Watts/170 Watts
or about 59%.  Dynamometer testing of the alternator shows it has an efficiency
of 85% at this condition which means the turbine is running at 69%.  Now 120
VAC is generated so no transformers are used at the generating site.  The same
transformer set used with the Delco installation is used at the battery end.
About 6 Amperes are delivered to the 12 volt battery.  This gives an overall
efficiency of 72/170 or 42% water to wire (water to battery?).
With this system appliances can be run directly off the alternator output as
long as this requirement is less than the available power.  This creates a
hybrid setup that produces both 120 VAC @ 60 Hz. and 12 VDC.  A future article
will discuss how to deal with more difficult sites.
