A creek's power comes down to two numbers: head (vertical drop in feet) and dry-month flow (gallons per minute). Measure head with a garden hose and flow with a bucket and stopwatch. Net head times flow divided by 10 gives watts. Weigh those watts against the full installed cost before you buy.
If you are sizing up your creek this week, after the fall rains, you are looking at the wrong creek. The water you see in November is not the water you will have in August, and August is the flow a turbine has to live on.
The test costs almost nothing. A garden hose, a funnel, a bucket, a stopwatch and a helper will tell you most of what a dealer would. Here is how to measure both numbers on a Saturday, turn them into watts, and decide honestly whether the creek is worth a turbine.
Head is the vertical drop from the intake, where water leaves the creek, down to the turbine. Only the vertical part counts. University of Wyoming Extension puts it plainly: head "is not a measure of the horizontal length of a pipeline," and a longer pipe can cost you energy through friction.
Flow is how much water moves past a point, here in gallons per minute (gpm). One cubic foot per second (cfs) is about 449 gpm.
Power is the product of the two. A big drop with a trickle makes little power, and so does a lazy creek with two feet of fall.
Gross head is the drop you measure on the ground. Net head is what is left after pipe friction takes its cut. The turbine only sees net head, so that is the number that goes in the math.
This is the U.S. Department of Energy method for small sites:
DOE advises subtracting an inch or two from each reading, since water can keep trickling after both ends are level: "It is best to be conservative." Four legs of about 10 ft gives roughly 40 ft of gross head.
Skip your phone. GPS elevation is too coarse for a 30 or 40 ft drop, and DOE says an altimeter only makes sense with several hundred feet of fall.
Yes, and on many farms it is faster. If a spring line, stock-water line or irrigation pipe already runs from the intake area toward the turbine site, put a gauge on the low end.
Read it static, with no water flowing, because a flowing reading already has friction loss baked in. Multiply psi by 2.31 to get feet of head. Wyoming's example: 40 psi x 2.31 = 92.4 ft. A gauge reading 17.3 psi works out to 40 ft. No pipe yet? A hose filled end to end and capped with a gauge does the same job.
Divert the whole flow with a board or a short length of pipe so all of it pours into a bucket, and time the fill. Gallons divided by seconds, times 60, is gpm. A 5-gallon bucket that fills in 10 seconds is 30 gpm. Time it three or four times and average.
Example: 4 ft wide by 0.5 ft deep is 2 square feet. The bottle covers 20 ft in 10 seconds, or 2 ft per second. 2 x 2 x 0.7 is 2.8 cfs, about 1,257 gpm. That raises the next question: how much of it can you take?
DOE's safety rule: this method "isn't recommended if the stream is fast-flowing and/or over your calves." Work from the bank or wait for lower water.
The USGS StreamStats service can outline the drainage above a point and estimate flow statistics at ungaged sites, though what it offers varies by state and it may not cover a small creek well. Your county engineer and NRCS office may also have numbers.
If a USGS gage sits on the larger stream your creek feeds, find it and learn to read it, then watch its seasonal swings in the Streamflow & Irrigation Watch. A river gage is not your creek's flow, but its yearly low tells you when to go take your own reading.
The lowest one. DOE: unless you build a storage reservoir, "you can use the lowest average flow of the year as the basis for your system's design." Wyoming adds that an average flow with big seasonal swings "may not be the most accurate measure" for sizing.
That is why a November reading misleads. Fall rain and spring runoff can make a creek look far bigger than it runs in late summer, and a turbine sized to that water sits starved for months. Measure now for a baseline, again at the late-summer low, and monthly for a year if you can before spending real money.
You also cannot take the whole creek. ATTRA, the sustainable agriculture program run by NCAT, warns that over-diverting or damming a stream can cause "long-lasting environmental damage, including reduced in-stream flow, increased water temperature, altered fish habitat." Plan on a share of the low flow, not all of it.
The DOE rule of thumb:
Net head (ft) x flow (gpm) / 10 = watts
DOE says it represents "a system with 50% to 70% efficiency or more," typical of micro-hydro. Wyoming uses the same rule for a turbine around 53 percent efficient. ATTRA divides by 12 instead, noting the divisor "may range from 9 for larger AC systems to 13 for smaller battery-based systems." Use 10 to plan and 12 to stress-test.
Worked example:
Hold on to that $120. It matters below.
How small is too small? ATTRA says head under 10 ft "is probably going to be uneconomical" and treats 10 gpm as the lower flow limit. Wyoming says 3 ft can be harnessed but sites generally need 10 ft or more to pay. Treat 10 ft and 10 gpm as floors, not targets.
The pipe down to the turbine, the penstock, is where good sites get ruined. We ran example sites through the Hazen-Williams formula for smooth PVC (C = 150) at nominal pipe sizes. Real inside diameters vary by schedule, so treat these as approximate.
| Site | Pipe size | Friction loss | Net head | Watts (/10 rule) |
|---|---|---|---|---|
| 40 ft head, 30 gpm, 400 ft of pipe | 1.5 in | 29.7 ft (74%) | 10.3 ft | ~31 W |
| Same site | 2 in | 7.3 ft (18%) | 32.7 ft | ~98 W |
| Same site | 3 in | 1.0 ft (3%) | 39.0 ft | ~117 W |
| 100 ft head, 60 gpm, 800 ft of pipe | 2 in | 53 ft (53%) | 47 ft | ~282 W |
| Same site | 3 in | 7.4 ft (7%) | 92.6 ft | ~556 W |
On the 100 ft site, one pipe size up nearly doubles the output from the same creek. On the 40 ft site, 1.5-inch pipe throws away three quarters of the head.
How much loss is acceptable? DOE suggests estimating 5 to 10 percent for preliminary numbers, while Wyoming notes friction "is often 20 percent of gross head." Budget 10 to 20 percent, and if your pipe loses more, price the next size up.
With head and low flow in hand, the free Micro-Hydro Power Planner does the next step. Enter head and flow (gpm or cfs) and it suggests a turbine class: Pelton for high head, Turgo or crossflow in the middle, propeller for very low head. It also estimates annual kWh and savings at your electric rate, with a chart of output from 25 to 125 percent of your flow.
Three settings keep the answer honest:
Its annual figure already assumes the system runs 90 percent of the time.
The best itemized costs we found are old. Wyoming Extension bulletin B-1285 reprints contractor-installed estimates for AC systems that ATTRA published in 2011:
| System | Site | Installed cost (2011 est.) | Cost per watt |
|---|---|---|---|
| 100 W | 16 ft head, 63 gpm | $15,050 | $150.50 |
| 400 W | 80 ft head, 63 gpm | $14,450 | $36.13 |
| 3.5 kW | 163 ft head, 222 gpm | $21,450 | $6.13 |
Those are not current prices, so get a quote. The lesson holds: intake, powerhouse and controller costs run several thousand dollars at any size, so cost per watt drops sharply as head rises. The 400-watt site cost less than the 100-watt site on the same flow because it had five times the drop.
Now go back to the worked example. About $120 a year against a five-figure system never pays back on the electric bill alone. ATTRA adds that if you still buy some grid power, "the 'meter charge' may eliminate any significant cost savings."
Micro-hydro earns its keep when:
A creek's edge over solar is that it keeps running at night and through cloudy weeks, as long as the water holds. Run the solar numbers for the same load and compare.
On incentives: as of November 1, 2026, USDA's REAP page lists hydropower as eligible for farms but says it "is not accepting REAP grant applications at this time." Guaranteed loan applications are still open. Ask your state USDA Rural Development energy coordinator.
Probably, so check before you buy. DOE: "Each state controls water rights; you may need a separate water right to produce power, even if you already have a water right for another use." Wyoming adds that water crossing your land "does not necessarily mean that you have the right to use it, even for a non-consumptive use such as hydropower." In much of the West, water rights do not automatically transfer with the land.
Start with your county engineer, then your state water agency and energy office, and ask whether the project needs a filing with FERC or the Army Corps of Engineers.
Measure head with a hose or gauge. Measure flow with a bucket or float, and again when the creek is lowest. Multiply net head by flow, divide by 10, and size the pipe so it does not steal the head you found.
If the watts cover a real need, especially one that would otherwise mean a line extension or a generator, get quotes. If they come to a hundred-odd dollars a year against a grid bill, you have saved yourself an expensive lesson, and the creek may still run a gravity-fed stock water system.
Multiply net head in feet by flow in gallons per minute, then divide by 10 to get watts. That rule from the U.S. Department of Energy assumes about 50 to 70 percent water-to-wire efficiency, which fits most small systems. For a cautious estimate, divide by 12. Use the flow from your driest month and the head left after pipe friction, not the raw drop.
As a rough floor, about 10 feet of head and 10 gallons per minute of flow. ATTRA calls heads under 10 feet probably uneconomical, and University of Wyoming Extension says sites generally need 10 feet or more to pay, though as little as 3 feet can be harnessed. Head pays: in old Wyoming estimates, an 80-foot site cost less to install than a 16-foot site at the same flow.
A typical small site makes well under a kilowatt. A creek with 40 feet of drop and 30 gallons per minute in late summer, fed through 400 feet of 2-inch pipe, makes about 98 watts around the clock. That is roughly 860 kWh a year, worth about $120 at 14 cents per kWh. Larger drops and flows scale up quickly.
Measure during the late-summer low, because that is the flow a turbine has to run on for months. A reading after fall rain or spring runoff can run far higher than the dry-season flow. Take a baseline reading now, measure again in August or September, and ideally measure monthly for a year before buying equipment.
It depends on what it replaces. Against a normal grid bill, small systems rarely pay back, because installed costs run well into five figures and a 100-watt site makes about $120 a year. Micro-hydro makes sense when the alternative is extending a power line, running a generator daily, or powering a load right next to the creek.
Join our list for practical guides on farm tech, precision agriculture, and tools that work.