Hydro · Intake, penstock and winter
The turbine gets the attention and the pipe decides the outcome. Friction rises steeply as diameter falls, so the cheaper penstock can quietly give away more head than the site can spare, every second, for thirty years.
Where the energy goes
Four stages sit between the water at your intake and the electricity at your panel, and each takes a percentage of what reaches it. Micro-hydro design guidance gives typical figures of 85 to 90 percent for the penstock, 80 to 90 percent for the turbine, 80 to 90 percent for the generator, and 85 to 90 percent for the transmission line, with overall efficiency being their product[1].
Take the midpoint of each and multiply: roughly 55 percent makes it through. Take the low end of each and it is 46 percent. That is why the screening estimate this section has used throughout assumes about half the energy in the water arrives as electricity.
The Department of Energy's shortcut implies a system efficiency near 53 percent. The stage-by-stage chain gives 55 percent at its midpoint. Those were arrived at independently, and they agree.
The useful part is knowing which stage to argue with. Turbine and generator efficiency are bought. Line efficiency is a wire gauge decision. Penstock efficiency is the one you control most directly and the one most often thrown away, which is what the rest of this page is about.
The intake
Everything upstream of the pipe exists to deliver water that the turbine can survive. The intake is where most maintenance actually happens, and a badly designed one turns a quiet system into a weekly errand up a hill.
A low weir raises the water enough to feed the offtake at low flow without impounding anything. It has to keep working when the stream drops in August and survive when the stream rises in April.
Leaves, needles, sticks and gravel all destroy runners or block nozzles. A coarse trash rack handles the large material, and something finer has to handle the rest before it reaches the pipe.
A forebay is a tank where velocity drops enough for sand to fall out before the water enters the penstock. It also keeps air out of the pipe when flow changes suddenly, and gives a place to draw down and clean.
Frequently a permit condition rather than a courtesy. Regulations commonly require that all diverted water be screened so fish cannot pass into the penstock, which sets a maximum slot size and a maximum approach velocity at the screen face.
Any screen fine enough to stop pine needles is fine enough to clog with them. The question worth asking of any intake design is not whether it screens well, but what happens to it during a week of leaf fall while you are away. An intake that needs manual clearing after every storm will eventually be abandoned, and the system with it.
The screen that cleans itself
A Coanda-effect screen is an inclined panel of wedge wire with the wires tilted so that fast-moving flow passing across the surface has thin layers sheared off the bottom of the water column and directed through the slots. Debris and fish stay in the sheet of water and carry on over the end of the panel[2].
The Bureau of Reclamation describes these screens as having large flow capacities and being hydraulically self-cleaning without moving parts, so they require minimal maintenance, which makes them well suited to remote sites without electrical power[2]. For a household system up a hill, that combination is close to the whole argument.
They are not free. A Coanda screen needs a drop across the panel to work, which spends some head, and it needs to be sized properly for the flow it will see. The Bureau publishes design guidance and a numerical model for computing screen capacity, and both are publicly available[2].
A conventional screen and a rake work perfectly well, provided somebody rakes them. Weigh the cost of a self-cleaning intake against a decade of walking up in the rain, and against what happens the one time nobody does. Blocked intakes do not merely stop production. They can starve a penstock and let air into a system that was designed to run full.
Sizing the pipe
The design rule is straightforward: size the pipe so that pipeline friction costs no more than 10 to 15 percent of gross head, and stay toward the low end of that range on low-head sites where there is less to give away.
What makes this consequential is how sharply friction responds to diameter. Halving the bore does not double the loss. It multiplies it many times over. Here is the running example from this section, worked at four pipe sizes.
150 gallons per minute, 400 feet of PVC penstock, 60 feet of gross head
2 inch: loses about 145 feet. The pipe cannot pass this flow at all. Water velocity above 15 feet per second.
3 inch: loses about 20 feet, a third of the site. Net head 40 feet, output near 600 watts.
4 inch: loses about 5 feet, roughly 8 percent. Net head 55 feet, output near 825 watts.
6 inch: loses under a foot. Net head 59 feet, output near 890 watts.
Going from 3 inch to 4 inch pipe raises output from about 600 watts to about 825, an increase near 38 percent, from the same stream and the same turbine. Going from 4 inch to 6 inch buys only another 8 percent for a considerably more expensive pipe.
That is the shape of the whole problem. There is a step where the pipe stops being the limiting factor, and the money spent before that step returns far more than the money spent after it. On this site the answer is 4 inch, and it is not close.
The 55 feet of net head that falls out of the 4 inch case is where the 54 foot figure used throughout this section came from. The 10 percent friction allowance the measuring guide applied as a rule of thumb turns out, for this pipe on this run, to have been about right.
Material and burial
The materials in common use for small and micro hydro penstocks are medium and high density polyethylene, rigid PVC, and mild steel, chosen for suitability, availability and affordability rather than for elegance.
Tough, weather resistant, smooth inside so friction is low, and forgiving of ground movement. Sizes up to about 75 millimeters can be coiled, which matters when the route is a footpath. Joining it properly takes fusion welding skill, so factor that in.
Inexpensive, widely available in a range of pressure ratings, and simple to join. It becomes brittle with age and ultraviolet exposure, so it belongs underground rather than lying on a hillside in the sun.
Handles the highest pressures and the roughest ground. Heavy, expensive, needs corrosion protection, and needs welding. Generally the answer only on high-head sites where plastic cannot take the static pressure.
Static pressure at the turbine is gross head times 0.433 pounds per square inch per foot. Sixty feet of head is about 26 psi, which almost any pipe handles. Three hundred feet is 130 psi, which rules a good deal of material out. Whatever the pipe, its rating has to exceed the static pressure with margin, because the transient pressure when a valve closes is higher still. That is the subject of the next guide.
A surface penstock has to be protected from falling trees, livestock and vehicles, and anchored with concrete at the points where the line changes direction, since a pipe full of moving water pushes hard at every bend. Burying it addresses those problems together.
It also addresses temperature, ultraviolet degradation of plastic, and the freezing discussed below. The trench is the largest labor item in most micro-hydro projects and it repays itself over the life of the system.
Winter
This is the reassuring part. Water moving continuously through a buried pipe carries heat with it and rarely freezes, which is why hydro is the one generation source in this section that gets better in winter rather than worse. Streams often run higher, and the load the household places on the system is at its peak.
The vulnerabilities are specific rather than general, and there are three of them.
In cold conditions, ice forms in turbulent open water as suspended crystals and as ice adhering to submerged surfaces, and both accumulate on screens. Screen performance in cold climates is enough of a distinct problem that it has its own research literature[3]. If you are in a cold region, raise this specifically with any screen supplier.
A running system protects itself. A stopped one, in a shallow pipe, in January, does not. This is why an outage during a cold snap becomes an urgent problem rather than an inconvenience, and why a system that shuts down when it detects a fault needs somebody to know it has.
Whatever is above ground: the last stretch into the powerhouse, the nozzle, the tailrace outlet. Spray freezes on everything near a working turbine, and a tailrace that ices over backs water into the machine.
Bury the penstock below your local frost depth, which your county extension office or building department can tell you. Put a drain valve at the low point so the line can be emptied deliberately if the system has to be taken out of service in winter. Keep the powerhouse above the tailrace so it cannot flood if the outlet ices. The system running normally is the easy case. Everything worth designing for is what happens when it is not.
Common mistakes
The example above gives away a third of the site to save one pipe size. The saving is once. The loss is every second for thirty years.
Friction depends on the bore. Subtract twice the wall thickness from the nominal size before calculating anything, or the pipe will be smaller than the one you designed.
Fish exclusion is often a permit condition with a specified maximum slot size and approach velocity. A screen that satisfies the turbine may not satisfy the regulator.
Cheaper to install and worse in every other way: ultraviolet degradation, freezing, livestock, falling limbs, and thrust at every bend needing anchoring.
Running the penstock straight off the intake sends sand to the runner and lets air into the pipe whenever the water level dips. The tank is not optional on a stream carrying sediment.
Elbows add turbulence losses and thrust that has to be anchored. A slightly longer run with a gentler line often delivers more head to the turbine.
Next
A pipe full of moving water carries momentum, and a turbine spinning under load carries rotational energy. Neither can be switched off the way a solar array can. Closing a valve quickly produces a pressure spike the pipe may not survive, and disconnecting the electrical load lets the machine run away.
The control guide covers how a hydro system is given somewhere safe to put power it is not using, which is where the electrical safety of the whole arrangement lives.
The machine the net head on this page decides, and why part-flow performance matters.
Read the guideWhere these components sit in the chain from stream to panel.
Read the guideGross head, net head, and the difference this page is entirely about.
Read the guideSources
Head loss figures in the pipe size comparison are calculated for PVC at 150 gallons per minute over a 400 foot run using standard friction relationships, and are illustrative of the relationship between diameter and loss rather than a design for any particular site. Actual loss depends on pipe roughness, internal diameter, fittings, bends and valves. Calculate your own run, and confirm pressure rating against static head with margin for transient pressures.