Greater Cincinnati Water Works The Great Tennyson Pump Station Mystery. Imagine the result

Transcription

1 Greater Cincinnati Water Works The Great Tennyson Pump Station Mystery Imagine the result

2 OTCO Workshop August 12, 2009 Authors: Gary L. Hoffman, P.E. ARCADIS Mark Ginty, P.E. Greater Cincinnati Water Works William Barhorst, P.E. ARCADIS John Anspach, P.E. John Anspach Consulting Special Thanks: Carel Vandermeyden, P.E. Chief Engineer Greater Cincinnati Water Works Imagine the result

3 Greater Cincinnati Water Works Regional water utility serving over 1,200,000 customers over 800 square miles Supplies over 50 billion gallons of water per year Two water treatment plants (220 mgd and 40 mgd) Over 3,100 miles of water mains 24 pumping stations 32 storage tanks

4 Tennyson Pump Station Components Flow from clearwell 10 feet diameter shaft, ~170 feet deep 8 feet diameter tunnel miles long, concrete lined, built in 1951 Stand pipe for surge protection at the end of the tunnel Tennyson pump station 160 to 185 mgd capacity 6 25 mgd, expandable to 8 pumps

5 Why Can t We Pump at Least 100 MGD Consistently? Vacuum priming problem? Pump packing problem? Pump impeller problem? Partially closed valve? Standpipe problem? Access shaft problem? Tunnel blockage? Is there a hydraulic issue?

6 Why Can t We Pump at Least 100 MGD Consistently? Tunnel Hydraulics Modeled TENNYSON TUNNEL HYDRAULICS CW#1 TO PUMP STATION SHAFT HEADLOSS VS. FLOW HEADLOSS (FT) Tunnel Calibration C = 125 (based on 2005 data) FLOW TO TENNYSON (MGD) 2004 data appears to have incorrect gage calibration or incorrect reference elevation.

7 Why Can t We Pump at Least 100 MGD Consistently? Tennyson Pump Station Layout?

8 Why Can t We Pump at Least 100 MGD Consistently? NPSH Checked Out OK TENNYSON P.S. - OPTIMAL CAPACITY ALL PUMPS NPSH VS. FLOW NPSH (available) Note: Existing pump curves indicate 16 feet NPSH required at 25 MGD duty point. NPSH (FT) NPSH required with 1.3 Factor of Safety NPSH (required) POTENTIAL CAVITATION ZONE FLOW TO TENNYSON PS (MGD) CALCULATED LIMIT OF PUMPING

9 Why Can t We Pump at Least 100 MGD Consistently? What About Surge? Measured on gauge 16.5 ft below pump centerline

10 Why Can t We Pump at Least 100 MGD Consistently? Hydraulic Transient Analysis Performed Candidates for Surge Problems Due to: 2.3 mile tunnel with no branches to absorb pressure waves One source Clear Well at Richard Miller WTP One Surge tank at the Pump Station open to atmosphere Observations of the Surge Well level dramatically changing during Pump on/off operation added to the hypothesis that surge may be the problem

11 Why Can t We Pump at Least 100 MGD Consistently? Start-up Modeled with Fourth Pump Pump CS-2 Start-up Flow Suction (mgd) (MSL ft.) steady 520 Pump CL Atm. Pressure 33.25

12 Why Can t We Pump at Least 100 MGD Consistently? NPSHA Still OK Under Surge Pump Tests Sept 2007 NPSH Available for Pump Starting Table 5 Tennyson Pump Station NPSH Available During Pump Starting Pump Clearwell HGL Suction HGL NPSHA Before Total Starting Before Start (ft) Before Start (ft) Start (ft) Surge (ft) Minimum NPSHA During Start (ft) 4th Pump (CS-2) 5th Pump (CS-3) 6th Pump* (CS-1) * -3.8* 28.9* -9* 23.7* - *projected from model

13 Narrowed Down to Three Pump Tests May 16 and 17 Purpose of the Pump Test on May 16 and 17, 2008 Was to Evaluate the Remaining Three Identified Possibilities for the Limited Capacity: - Starting Procedure? - Pump Wear? - Air Intrusion?

14 Narrowed Down to Three Pump Tests May 16 Shut-off Head Test Design Actual Pump PSI PSI KW EH EH EH-3 * CS CS CS *est. Shut-off Head Testing Performed to: Determine if adequate pressure available Gauge pump wear Manual Prime and Air Relief, Results showed: Adequate Pressure Minimal Wear

15 Narrowed Down to Three Pump Tests May 17 Attempt to Bring All 6 Pumps Online under Two Scenarios Scenario 1 Start EH Pumps under normal conditions and bring CS pumps online by manually isolating all the prime to the pump and bleed air from casing; keep cone check closed until adequate pressure is developed and then open valve. Scenario Two Have EH Pumps running; bring CS pumps on during normal operation (i.e., normal priming condition and no air relief from casing) except cone check valve held closed until adequate pressure reached.

16 Narrowed Down to Three Pump Tests May 17 Scenario 1 Scenario 1 Controlled Valves, manual prime and air relief, started CS-1, CS-2 and CS-3 in that order. Success! MGD OBSERVATIONS EH MGD EH MGD EH MGD CS MGD CS MGD CS MGD -Starting Pumps CS-2 and CS-3 was noisy and pressures on suction side oscillated before steady-state -CS-3 was the last pump turned on and discharge pressure gauge dropped as low as 20 psi before stabilizing at 84 psi and opening valve

17 Narrowed Down to Three Pump Tests May 17 Scenario 1 Shut-off Head Test May 16 May 17 May 17 Pump KW KW Suction below CL CS ft. CS ft. KW directly correlates to specific gravity of liquid pumped. CS-2 and CS-3 pumping something other than water actually a water and air mixture. Air is getting into the casing. CS ft.

18 Narrowed Down to Three Pump Tests May 17 Scenario 2 Scenario 2 Controlled Valves, but normal priming, no air relief, started CS-3, attempted CS-1 and CS-2 in that order. With CS-3 running, CS-1 was started against a closed valve and after 4 minutes of pumping, did not develop enough pressure to overcome the discharge pressure and was then shutdown. The test was halted with the three EH Pumps running and CS-3 running. Result: only four of six pumps were able to be started under this scenario.

19 Narrowed Down to Three Pump Tests May 17 Why is air in the pump casing? Priming system? Packing? Coming out of solution?

20 Narrowed Down to Three Pump Tests May 17 How do the hydraulic conditions affect air volume in pump? Pump Starting Clearwell HGL Before Start (ft) Suction HGL Before Start (ft) 4th Pump (CS-2) 5th Pump (CS-3) 6th Pump* (CS-1) 7th Pump* (Prop. EHS-4) * -3.8* 6.0* -5.0*

21 Narrowed Down to Three Pump Casing Full of Water Ideal conditions usually when no more than three pumps running and air relieved.

22 Narrowed Down to Three Pump Casing Half Water and Half Air Even with priming system, not likely to be better than this with 4 pumps running.

23 Narrowed Down to Three Pump Casing Full of Air With 4 or more pumps running, suction HGL below pump centerline, is priming system able to pull water into casing or is it a straw in the ocean?

24 Narrowed Down to Three Air is the Problem What s next? Review priming set-up and capacity Evaluate packing Recommend automatic air relief on top of casing Evaluate pump start-up sequence and valve opening timing

25 Improvements In Progress Original Prime Location New Prime Location

26 Thank You