Simplified current meter measurements - NTNU · Simplified current meter measurements used to verify improvement of turbine efficiency after runner replacement International Group

Simplified current meter measurements used to verify improvement of turbine efficiency after runner replacement

International Group of Hydraulic Efficiency Measurements Trondheim 2012 Leif Parr and Erik Nilsen, Norconsult Jan Øystein Rafoss, E-Co Petter Thorvald Krogh Østby, Rainpower

About the authors

• Leif Parr and Erik Nilsen, Norconsult Norway • Hydraulic efficiency tests since IGHEM 2010:

- 36 thermodynamic measurements - 1 pressure-time test, 4 current-meter tests

• Jan Øystein Rafoss - E-Co • The 2nd largest producer of electric power in Norway • 9.7 TWh/year, 2800 MW

• Petter Thorvald Krogh Østby - Rainpower

Contents

• Solbergfoss I - plant description • Previous current-meter tests • Current-meter test setup • Uncertainty • Design of new replacement runner • Guarantee measurements of replacement runner • Conclusions

Solbergfoss I - plant description

Solbergfoss I - plant description

• Solbergfoss I • Run-of-the river plant in Glomma • 13 Francis units of 8.5 MW - 12.5 MW • Head 11 - 21.5 m

• Solbergfoss II • in parallell to Solbergfoss I • 1 Kaplan unit of 100 MW

• This presentation concerns unit 9 in Solbergfoss I • P = 8.5 MW, Hn = 21m, n = 150 rpm

Previous current-meter tests

• Current - meter tests of all units during commissioning • Unit 9 in 1925 - no documentation available • Unit 9 in 1957 - documentation of results and velocities • Unit 10 in 1957 - report by Neyrpic Grenoble with uncertainty

analysis

Solbergfoss I at night (Photo: Tormod Hanstad)

Current-meter test setup

measurement section

Current-meter test setup

• Setup full current-meter test • Array of 77 points • Horizontal beam traversed in

9 positions • Position of simplified test in

red

Current-meter test setup

• Nilsen - 1997 - Analysis of velocity profiles from 1957 test • 3rd position from bottom chosen

- ease of installation - regular velocity profile for complete flow range

• Little loss in accuracy in reduction from 9 to 7 current-meters • Q = kv7 · A · average(v1 - v7) • kv7 = 0.9625, stdev 0.01

Q v_av. v_av. Velocity v_av. Velocity1 2 3 4 5 6 7 8 9 (m3/s) (m/s) (m/s) coeff. (m/s) coeff.

0.96 0.95 1.01 1.06 1.00 0.99 0.99 0.95 0.85 30.55 0.953 0.973 0.979 0.993 0.9591.13 1.24 1.16 1.25 1.19 1.19 1.19 1.16 0.99 36.68 1.144 1.167 0.981 1.196 0.9561.26 1.29 1.30 1.39 1.30 1.30 1.31 1.21 1.10 41.00 1.279 1.272 1.005 1.299 0.9851.29 1.23 1.39 1.45 1.31 1.35 1.36 1.26 1.11 41.40 1.291 1.306 0.988 1.336 0.9661.36 1.32 1.48 1.54 1.42 1.53 1.44 1.36 1.16 44.05 1.374 1.400 0.981 1.440 0.9541.34 1.37 1.53 1.56 1.46 1.48 1.49 1.43 1.22 45.75 1.427 1.431 0.997 1.475 0.9671.55 1.53 1.59 1.68 1.55 1.57 1.57 1.52 1.29 47.95 1.495 1.536 0.974 1.570 0.9521.24 1.18 1.39 1.43 1.34 1.37 1.37 1.31 1.12 41.58 1.297 1.305 0.994 1.341 0.9671.18 1.18 1.34 1.41 1.29 1.37 1.32 1.26 1.09 40.15 1.252 1.271 0.985 1.311 0.955

Average 0.9871 Average 0.9625St.dev. 0.0100 St.dev. 0.0102

9 CURRENT METERS 7 CURRENT METERSVelocity (m/s) in pos no.:

Current-meter test setup

Current-meter test setup

Uncertainty

• 1957 test of unit 10 - total uncertainty in flow ±1% • Same test setup as unit 9 - similar uncertainty

• Simplified test setup - uncertainty in velocity coefficient kv7 • Standard deviation 1%, 95% confidence -> ±2%

• Simplified test setup - uncertainty in flow ±2.2%

Design of new replacement runner

• Low head and high discharge, better suited for Kaplan • Large head variation (Hmax / Hmin = 2) • Design differs significantly from old runner • No reference model turbine • 50 designs run through

CFD • Hub, band and blades of

carbon steel, painted

Guarantee measurements of replacement runner

Guarantee measurements of replacement runner

• 0.01 percentage points deviation between repeated points • Net increase in turbine efficiency 3.2 percentage points (peak) • Peak efficiency shifted 7.1 MW -> 8.9 MW

• Winter-Kennedy calibrated with flow within ±1% in efficiency

• Turbine power at peak is higher than calculated

• Replacement runner contract guarantee is verified

Conclusions

• Reduced test time - 1957 test time 10 days - 2012 test time 2.5 days - WK test time hours

• Simplified test setup accepted by supplier of replacement runner

• New runner: Peak efficiency 3.2 percentage points increase, 18% increase in power

• Location of peak efficiency deviates slightly from calculations