Follow up on Risky Hydro thread...
 
...and the question of using negative pressure in a pelton
turbine housing to augment power output.
 
This past week we got our home made ('shop made') 14kw pelton
turbine back together and running after rebuilding the runner
when last years' model proved to be too lightly constructed.
The new runner, estimated as 7 times stronger at the bucket
attachment point that was the trouble before, seems plenty
sturdy.
 
This time we have provided three 6" outlets for the turbine
discharge water (two was not enough).  The outlets have 6"
corrugated drain tubing leading down about 20 feet of vertical to
the stream bed.  The ends are submerged in half a 55 gal drum
kept full by the flow in the stream.
 
During operation, air entrained with the discharge water does
cause a slight negative pressure to develop in the housing.  The
vacuum is slight , however, and will not register on a gage
calibrated 0-30"Hg vacuum, but can be felt by removing the gage
and placing a finger over the hole.  Measuring by water gage I
got only 2" water column.  
 
In an effort to get the effect started, I hooked up a jet pump to
aid in evacuating the housing.  This produced only about 2' of
vacuum on the water gage.  I think we must not have the housing
sealed well enough.  Perhaps the shaft seals are the problem.  
 
Otherwise I'm real pleased with our 'home built' turbine.
 
Spec's. :
 
Head: 470'  Flow: up to 300gpm (pipeline limits flow)  
 
Water source:  18,000,000 gal reservoir formerly used as an
industrial water supply.  Max. depth 20'. 
 
Watershed: about 70 acres (.. wish it were more!)
 
Pipeline: 1840' of 4" HDPE (Black poly) heat welded in 40'  
lengths in staged pressure ratings from 110psi to 200psi.
 
Generator:  20hp, 1750rpm, 3ph, 575V, TEFC Toshiba Premium
efficiency motor used as an induction generator at 480V system
voltage 
 
Wire run:  about 2000' of #8AWG Tray Cable strung on insulators
attached to trees.  (Up to 60V drop in transmission = 1.5 kw loss
at max. power)
 
Capacitors:  7 Kvar at generator end of transmission line to
correct power factor and minimize amps and losses in the wire
run.
 
Grid Intertie:  Power Co. required  Over & Under Voltage, and
Over & Under Frequency relays.  We used inexpensive devices from
TimeMark Corp. and got off cheep at about $350 total for the two.
This system is on an industrial property served by a single
meter.  The Mill is now shut down and dismantled on that site,
but there are 10 occupied Company owned homes and some support
facilities remaining.  Since the average power use at the site
exceeds the capacity of the hydro system by several times, (and
probably because we are a good customer at the New Mill site) the
power Co. did not require any special metering, and just lets us
run the hydro downstream on the electrical system to partially
offset our use.
 
Turbine:  Designed (by me) and built-in house in our Co. shop. 
Two nozzle (at 90 deg spread, end and top) Pelton type. Runner is
24 bucket 9" pitch dia. w/ 1-3/8" shaft.  Housing is a welded
steel box with 1" thick plexiglass side panels and a removable
(for nozzle access) plexiglass partial top panel.  Nozzles are
2.5" Std. Pipe fittings flanged to the box (steerable) and
reduced at the tip to 3/4 NPT to take interchangeable brass
nozzles w/ various bore sizes.  The pipe fittings are faired w/
'liquid metal' internally at the reducer.
 
Runner:  Bolted-on bucket construction. Hub is machined from 3/4
Stainless steel w/ 24 pairs of 3/32" deep x 3/8" wide x 1-1/2"
long 'keyways' machined into the rim radially to accept the
bucket tangs.  The buckets are made of a castable plastic molded
around two 3/8" SS square bar pins (tangs) connected by 3/4" x
1/8" SS flatbar welded to the pins.  The pins are spaced 9/16"
and straddle the rim, like a riders legs straddling a horse,
sliding into the keyways for a tight fit, and pinned at the
'rider's ankles' with a machine screw. 
 
Bucket Pattern: Was taken loosely from detailed drawings in "The
Standard Handbook for Mechanical Engineers".  The pattern was
carved first in paraffin wax and then transferred by an
impression into autobody putty. A durable 'rough' was cast in
plastic from the bodyputty mold which I  sanded and shaped on a
grinder to make a finished pattern.  The bucket is 2-3/4 wide x
2" long and will accept a jet up to about 23/32" dia. 
 
Performance:  The turbine is running today at about 220 gpm w/ a
1/2" and a 9/16 nozzle and w/ a dynamic pressure of about 175psi
on the gage.  We are getting about 12.4 amps @ 550V w/ nearly
100% power factor or about 11.8 kw at the turbine shed.  At 100%
efficiency, 16.86kw would result, so we have about 73.5 %
efficiency, "water to wire" using those numbers.  The usual
expectation from purchased commercial equipment is about 75%.  So
we are pretty close to commercial equipment efficiency and I
wouldn't trust our pressure gage and volt and ammeter accuracy to
make a call that it is better or worse.
 
We have a manual reset arrangement to turn off the water in the
event of an electrical fault or grid outage.  This is advisable
so as not to waste water or overspeed the motor and bearings. 
This shutdown arrangement is a real Rube Goldburg.  A solenoid
keeps a latch closed until a fault condition.  On a fault , a
cable is released allowing a weight to drop which pulls on
another line closing the ball valves feeding the turbine.  To
prevent rapid closing of the valves and possible water-hammer
trouble, the weight is restrained by another cable connected to a
hydraulic door closer used as a dashpot.
 
We spent about $7,000 on pipe, wire, motor and other items, but
much of the electrical stuff, starters and meters, was scrounged
or salvaged from storage.  Depending on yearly rainfall, I expect
a payback on materials purchased in 2 to 3 years assuming 8
months operation per year (It would freeze in our -20 deg. F
winters) with average output of 9 KW and our incremental cost of
$.06/KWH.
 
I've had a lot of fun with this!
--- FLAME v1.1
---------------