Mechanical and Electrical Design of Pumping Stations - 11...

Chapter - 11 : Pump and Station Hydraulic Tests...

11-1. General...

Two types of tests are generally performed in connection with a pumping station and its pumping equipment before the station is built. Tests are run on pumps, either full size or model, to determine their performance and to demonstrate that the performance of the pump complies with specification requirements. Physical hydraulic model tests of the pumping station substructure may also be conducted to assess its hydraulic performance.

11-2. Pump Tests...

( a ) General.

The pump performance tests are conducted at the pump manufacturer’s facility. Field testing to prove performance at the completed station is more difficult and costly if not impossible in many cases. All pumps should be factory tested to determine their capacity, total head, efficiency, and horsepower requirements. Pumps normally should also have a cavitation test performed. This test is usually performed on a model. Model testing specifications and parameters can be found in Guide Specification CW 15160, Vertical Pumps, Axial and Mixed Flow Impeller Type.

( b ) Description.

Factory pump tests are either performed on full-size pumps or performed on a model pump of the full-size pump. If a model pump is used, it should be geometrically similar to that of the full-size pump and of the same specific speed. Either type of test is acceptable to check the ratings of the pump; however, because of size limitations, most manufacturers limit fullsize tests to pumps of less than 2 m3/s (75 cfs) capacity. A pump test consists of determining the total head, efficiency, and brake horsepower for a range of capacities. All testing should be witnessed by the District office design personnel performing the station design. The dimensions of the model and prototype impellers should be made by using drawings, measurements, and scaling factors. The performance factors measured include capacity, pressure head, horsepower, and suction pressure when cavitation performance is to be determined.

( c ) Performance tests.

All pumps for flood-control pumping stations should have their performance verified by tests. For installation of identical pumps in a station, only one of the pumps needs to be tested. Tests on similar pumps used for another station will not be acceptable as equivalent tests. The test setup should permit, and the specifications require, the pump to be tested over a range of heads starting at least 600 millimeters (2 feet) greater than the highest total head requirement or at shutoff and extending down to the lowest head permitted by the test setup. The test should, if an unstable range ("dog leg" in the head curve) exists. Sufficient test points should be run to adequately define the unstable range. This allows the pump manufacturer to demonstrate that their pump does not operate in the unstable range. The lowest head tested should be at least equal to the total head that occurs for 95 percent of the operation time during low head pumping conditions. For pumps with capacities greater than 11 m³/s (400 cfs), the model tests should be required to cover the complete head range required by the specifications including down to the lowest total head specified. All performance tests should be run at the same head at which the pumps will operate during actual duty. The readings of capacity and brake horsepower along with the total head will be used to determine the pump efficiency. For model tests, no correction factor for efficiency due to size differences will be allowed. Tests will be performed at water levels similar to that which will occur during actual operating conditions. An actual scale model of the station’s inlet and discharge systems is not warranted except for pumps over 14 m³/s (500 cfs). This requirement should also be used when the sump is not designed by the Government and is a part of the pump contract or has some complicated flow passage which has not had a sump model test. The pump test is used to ascertain the performance of the pumps, not how it reacts in the prototype sump except in the cases listed below. It is expected that the factory sump would be free of vortexes and adverse flows so that good results are obtained. Manufacturers are responsible for furnishing a pump that conforms to the specifications and meets the performance in the sump to be provided by the Government. The pump manufacturer should be held responsible for poor sump design, evidenced by vortexing and bad flow conditions within the sump, when the contract specifications require the sump to be designed by the pump manufacturer. Except for this special case, the pump manufacturer warrants performance of the pumps only, not the sump, and the activity within the sump would be the responsibility of the Government. Duplicate model pump sumps should include the sump from the inside of the trashrack. Any pump using a formed suction intake should be tested with this formed suction intake. Vertical pumps should be tested only in the vertical position.

( d ) Cavitation tests.

Cavitation tests are performed to indicate the operating conditions in which the pump will start cavitating. For purposes of design, it is assumed that cavitation starts when the pump performance starts to decrease as the effective sump level is reduced. The inception of cavitation definition has not been agreed upon by all the pump suppliers and users. A typical pump test consists of operating the pump at a fixed capacity while reducing the pressure on the suction side of the pump. As the suction pressure is reduced, a point is reached where a plot of the head-capacity curves deviates from a straight line. The Corps specifies the start of cavitation at a point where the curve starts to deviate from the straight line. Others use as the start of cavitation, a point where a 1- or 3-percent deviation in performance from the straight line occurs. Submergence requirements, as used in this manual, are based on the Corps criterion of zero deviations from the straight line portion. In most cases, some cavitation has already started at either point; therefore, a design allowance of extra submergence should be provided in addition to that indicated by the tests results. The submergence allowance is based on the estimated number of operating hours expected annually. The amounts of allowance are indicated in Appendix B. In all cases, the cavitation tests should be performed in a test setup that uses a variation of water levels on the suction side of the pump.

11-3. Station Tests...

( a ) General.

Hydraulic model tests of pumping station sumps and discharge systems should be performed by WES for stations with unique or unusual layouts. The procedure in ER 1110-2-1403 should be followed when requesting model tests. A decision should be made on the requirement for model testing during the General Design Memorandum stage so that the results of any testing are available during the design of the station. Test results are usually not available until 6 to 9 months after forwarding a work order to the test agency.

( b ) Sump model tests.

The primary purpose for performing a model test of a pumping station sump is to develop a sump design that is free of adverse flow distribution to the pump. Optimal flow into a pump impeller should be uniform without any swirl and have a steady, evenly distributed flow across the impeller entrance. However, it is usually not possible to obtain the optimal flow conditions without considerable added expense. Acceptable pump operation will occur when a deviation in the ratio of the average measured velocity to the average computed velocity is 10 percent or less and when the swirl angle is 3 degrees or less. Swirl in the pump column is indicated by a vortimeter (free wheeling propeller with zero pitch blade) located inside the column. Swirl angle is defined as the arc tangent of the ratio of the blade speed at the tip of the vortimeter blade to the average velocity for the cross section of the pump column. There should not be any vortex formations allowing entrance of air into the pump. In order to accurately simulate the field conditions, the model should include sufficient distance upstream of the station to a location where changes in geometry will not affect flow conditions in the sump. The prototype-to-model ratio is usually determined by the testing agency, but it should not be so large that adverse conditions cannot be readily observed. Normally the model should be sized to ensure that the Reynold’s number in the model pump column is equal to or exceeds a value of 1 × 10⁵. Reynold’s number is defined by the following equation:



( c ) Discharge model tests.

(1) General. These tests are performed to evaluate the performance of a discharge system. Usually two types of systems are investigated, discharges which form a siphon and/or through the protection discharges for large stations where the friction head loss would be a substantial portion of the total head of the pump.
(2) Siphon tests. The siphon tests are run to determine that a siphon will prime the system in the required time. This test is recommended when the down leg of the siphon system is long or it contains irregular flow lines and for pumps of 20 m³/s (750 cfs) or greater having a siphon built into the station structure.
(3) Discharge tests. A head loss test should be considered only for pump discharges with capacities of 20 m³/s (750 cfs) or greater and where the accuracy deviation for estimating the total head exceeds 20 percent of the total head. Other considerations would be the sizing of the pump and its driver. In some cases, a safety factor of 10 to 20 percent of the total head may not change the pump unit selection, and therefore the expense of a discharge test may not be warranted. For those stations where the size of the driver is close to its rating, a test may be in order to ensure that the driver would not be overloaded due to error in head determination.
(4) Other tests. Additional tests may also be required to fully prove the station acceptable. These could include energy dissipation tests of a saxophone outlet and a stilling basin or apron. Tests can also be made on models of existing pumping station sumps where operating difficulties have been experienced.