Kansas Geological Survey, Open-File Rept. 90-27
Annual Report, FY89--Appendix D
Use "Back" on your browser to return to the page you were viewing.
This study was initiated to investigate one aspect of the Dakota aquifer: the energy needed to pump water from high-capacity wells completed in the Dakota or a combination of the Dakota and the High Plains aquifers. High capacity for irrigation in Kansas is defined to be at least 800 gallons/minute (gpm) and wells of 1000 gpm are not uncommon. Irrigation of field crops in Kansas requires high rates of pumping and large amounts of water. Public water-supply wells by contrast are frequently of much smaller capacity. They may produce as much or more water but do so because they are pumped more hours than the average irrigation well. High capacity for this study was defined as 100 gpm or more.
The energy needs for pumping water in Kansas are not a trivial matter. Energy for pumping is one of the major variable costs in irrigated-crop production. In Kansas, there are about 3.5 million acres of irrigated land (Thomas, 1982) which use an estimated 5.12 million acre-feet of ground water. Power for all this pumping comes from natural gas engines, electric motors, diesel engines, and propane or liquid petroleum (LP).
The depth to water, pumping depth, pressure developed, the type and efficiency of the pumping plant, the value of the crops produced and the cost of the energy all affect the amount of energy use. High energy use will be tolerated if the price of crops is high relative to the cost of energy. Unfortunately, the factors that contribute to the irrigation costs are highly variable over the state. About 59% of all irrigation pumping in Kansas is done with natural gas. The major gas fields are located in the southwestern portion of the state where the largest acreage of irrigated crops are produced and is the cheapest form of energy where it is available. The cost, however, is not uniform. Well-head gas contracts may provide gas at $0.50/mcf (1000 cubic feet) while new gas to the same producer may cost in excess of $5.00/mcf. Electricity for irrigation pumping is next in popularity with 19% of the total. The cost varies statewide from about $0.08 to $0.12 per Kwh (Kilowatt-hour) but does not tend to vary as greatly within a given locale. Connection or horsepower charges, however, do vary greatly both in amount and method of computation. Some utilities charge a flat fee per horsepower per month all year long. Others charge per month but only during the irrigation season, while still others charge by the year or have no connection charges at all for irrigators. Many irrigators with lower power usage pay more for connection charges than actual power use. Electrical rates are controlled by the Kansas Corporation Commission, but the suppliers are allowed to charge different amounts depending upon how they class irrigation users. Diesel fuel and propane tend to be more uniformly priced over the state and the price per gallon tends to reflect the amount of heat energy per gallon.
Deeper pump settings, higher pressures, lower pumping plant efficiencies and larger volumes of water pumped are the primary factors that contribute to higher energy use. Factors such as the column pipe size and length, and the type and size of the pump drive line and pump head all contribute to the energy loss but normally are only minor.
The pump setting in the well is not often viewed as a variable but should be. The depth to water in the well is a function of inflow versus versus outflow. If recharge to the aquifer exceeds the pumping plus other losses from the aquifer, the water levels will rise. The pumping water level is also affected by the pumping rates and the transmissivity of the aquifer, gravel pack and well-screen system. Normally, little can be done about recharge rates or the natural properties of the aquifer, gravel pack and well-screen system. Well-spacing requirements and/or limiting pumping amounts and rates can and are being used to stabilize or control water levels in the region but these are generally beyond the control of the individual well owner. The owner, however, can reduce the pumping depth for a given rate of pumping by reducing the head losses into the well through proper design.
Highly efficient wells are the result of proper design and construction practices using geologic and test-pumping data. The size of the borehole and casing; the type, size and length of screen, the position of the screen; the gravel pack size; gradation; and the type and placement of grout seals and anything else that may effect the well performance are selected to efficiently deliver water to the surface. Unfortunately, too little of this is done for the ordinary irrigation well. In Kansas, the normal irrigation well-case size is 16 inches, well bores vary narrowly from 24 to 30 inches and well openings and gravel packs are frequently universal with a particular driller. The well openings start in the first portion of the borehole that exhibits a reasonable yield and extend to near the bottom of the hole. Gravel packs start at the required grouting depth and extend to the bottom of the well. Some drillers do not even test pump wells prior to installing the pump.
The amount of energy used to pump irrigation water in Kansas is unknown. No records are kept that would allow a direct determination to be made. Sloggett (1980) reported a total cost for pumping irrigation water in Kansas of $121 million which even to 1980 seems low. The author estimates that the energy expenditure is about 2.32 x 109 kwh at a total cost of about $204 million in 1989. Even allowing for inflation, the difference between these estimates seems too large. The acreage irrigated, the distribution of pump types and even the costs for energy for both estimates are quite similar. Sloggett used a total water use of 20.4 inch/acre and the author used 17.5 inch/acre. The biggest difference was in the cost of electricity. Sloggett used $0.55/kwh and the author used $0.88/kwh which is 60% higher, but the author's water use was 16% lower. Dividing the total costs by the total water use results in $1.75/acre-inch using Sloggett's figures and $3.32/acre-inch using the author's. Both of these values seem conservative, but the total cost in either case is substantial. Irrigation accounts for about 80% of the total water diverted and no estimate has been made for the cost of pumping the remaining 20%. The total cost for pumping water in Kansas is therefore greater than either of the estimates shown above.
The energy needs for pumping from a given site with known conditions can be easily calculated. Wells and pumping have been thoroughly investigated. Such calculations, however, do not reveal the general situation in the field. As noted above, there are many variables and how these are combined in field is the real question.
The southwestern group of eight counties contained 266 well records (56%), the west-central group of nine counties contained 112 well records (23%), and the north-central group of five counties contained 98 well records (21%). Twenty groups of five wells each were selected in proportion to the number of wells in the area: 12 groups or 60 wells in the southwest, four groups or 20 wells in the west-central, and four groups or 20 more wells in the north-central area.
Whether the well was founded entirely in the Dakota aquifer or was in an area where the High Plains and Dakota aquifers are interconnected could not be easily determined. Interconnection only exists in the southwestern portion of the state so a larger number of wells were selected for testing in the southwest than in the other two areas.
The pumping-plant test procedure included measurement of the static and pumping water levels, flow, water pressure at the pump outlet, pump and engine speeds and the energy or fuel input. Static water levels were measured with an electric probe or an air line. Discharge was measured with a Collins meter in the majority of cases, but a venturi by-pass meter, an eagle eye meter and a variety of propeller meters were also used. In general, the owner's meter was not used as a primary measurement but in several cases there was no reasonable alternative. The Collins meter is a velocity probe which is used to scan the velocity profile in the outlet pipe and obtain an average value. Normally, 12 velocity measurements were taken to obtain an average value and two centerline measurements were taken for checks on the equipment and procedure. Water pressures were measured with a variety of pressure gauges, and pump and engine speed with an electronic tachometer.
Energy or fuel input required a variety of measurements. Wherever possibe, electrical energy input was measured with the watt-hour meter that served the pump. Revolutions of the disc drive over at least 100 seconds were counted and the rate of power use computed. Measurements of current and voltage were also routinely made and the energy use was computed with these data where watt-hour meter readings could not be obtained. Power factors needed for these calculations were selected from data supplied by the Western Land Roller Company for U.S. hollow-shaft motors. When both sets of data were obtained, a power factor was computed from the voltage and amperage as a check on the watt-hour meter.
Diesel and Propane engine input was measured in pounds of fuel per unit of time using a portable scale and fuel tank. The rate of weight loss from the scale was measured during the test and the values converted to gallons. In the case of diesel fuel, the specific gravity was also measured in the field using a calibrated 500 mL volumetric flask and a triple beam balance. The temperature was measured and the specific gravity corrected to 60°F temperature.
Natural gas was measured with a bellows type meter. The owner's meter was used when conveniently located but the bellows meter was normally used and inserted into the gas line near the engine. The inlet pressure to the gas meter and the barometric pressure were measured and a correction factor was calculated for each gas-meter reading. Measurements were taken by counting the number of revolutions of the lowest volume dial on the meter for an interval not less than 100 seconds.
The well records provided data on construction and frequently on test pumping. The pumping tests reported in the well records were generally of longer duration than the pumping plant performance tests and should therefore produce better data values of specific capacity than the values calculated from the pumping tests.
Water temperature and pH measurements were made at each site and a water sample was taken for analysis of selected inorganic constituents. The pH measurements were made with a pocket-type meter of dubious accuracy, but it was the only equipment available at the time. The water analysis was conducted by the soils test laboratory, Department of Agronomy, Kansas State University.
Of the 476 well records provided, 100 were selected for pumping plant performance tests and 81 tests were performed. The grouping of counties, number of records per county, number of tests per county, the percent of the well records in each group of counties and the percent of tests conducted in each group are all shown on Table D.1. Figure D.1 shows the location of each area and the number of records and test per county. A complete compilation of data for the 81 test wells is listed in Appendix D.A and a brief summary is shown in Table D.2. The median, average and range of values for each parameter are shown to provide an easy comparison of the conditions found. In most cases, the range in values is so large and the number of samples so small that no further analysis of the data was perfomed.
Table D.1. Well logs provided
| Counties | Records | Tests | |
|---|---|---|---|
| I. Southwest | 1. Finney | 12 | 2 |
| 2. Grant | 64 | 6 | |
| 3. Gray | 8 | 0 | |
| 4. Hamilton | 33 | 7 | |
| 5. Haskell | 22 | 0 | |
| 6. Kearny | 38 | 12 | |
| 7. Morton | 27 | 3 | |
| 8. Stanton | 62 | 9 | |
| Subtotal | 266 (56%) | 39 (15%) | |
| II. West-central | 1. Barton | 1 | 0 |
| 2. Edwards | 2 | 0 | |
| 3. Ellsworth | 6 | 0 | |
| 4. Ford | 44 | 9 | |
| 5. Hodgeman | 28 | 9 | |
| 6. Ness | 6 | 0 | |
| 7. Pawnee | 13 | 2 | |
| 8. Rice | 6 | 0 | |
| 9. Rush | 6 | 0 | |
| Subtotal | 112 (23%) | 20 (18%) | |
| III. North-central | 1. Cloud | 24 | 5 |
| 2. Lincoln | 5 | 0 | |
| 3. Ottawa | 25 | 10 | |
| 4. Republic | 2 | 0 | |
| 5. Washington | 42 | 7 | |
| Subtotal | 98 (21%) | 22 (22%) | |
| Totals | 476 (100%) | 81 (17%) |
Figure D.1. Distribution of wells used in the energy-use study.
Table D.2. Well data summary by area.
| Area | Well depth (ft) |
Static level (ft) |
Specific capacity (gpm/ft) |
Age of pumping plant (yrs) |
|
|---|---|---|---|---|---|
| 1. Southwest | 525 | 167 | 17.2 | 11 | |
| 513 | 174 | 21.7 | 11 | ||
| (299-740) | (90-255) | (1.3-126) | (3-21) | ||
| a. Dakota | 510 | 163 | 5.9 | 10 | |
| 525 | 158 | 6.2 | 10 | ||
| b. Dakota-High Plains | 528 | 177 | 19.0 | 12 | |
| 509 | 180 | 26.2 | 11 | ||
| (299-740) | (95-255) | (5.9-126) | (4-21) | ||
| 2. West-central | 336 | 44 | 12.9 | 10 | |
| 284 | 59 | 14.9 | 9 | ||
| (107-577) | (20-167) | (1.0-60) | (2-15) | ||
| 3. North-central | 149 | 36 | 9.2 | 12 | |
| 151 | 46 | 15.7 | 11 | ||
| (75-244) | (20-87) | (3.4-60.0) | (5-15) | ||
As anticipated, the depth of wells and the static water level are greatest in the southwest and become progressively less to the east and north. The specific capacities, however, vary widely. The lowest values are for the Dakota wells of the southwest area, but all of the Dakota wells are lower on average than the Dakota-High Plains wells. Unfortunately, the range in values has too much overlap to allow wells to be identified by specific capacity alone.
The average age of the wells in each group is very nearly the same, which would indicate that drilling activities in the various areas has been about the same over the years represented by this group of records. Obviously, more wells were drilled in the southwest, but the average age of the group is 10 years and the subgroups vary by only one year.
Pumping Tests
The data from the pumping-plant tests are listed in Appendix D.B and a summary of the performance ratings are shown in Table D.3. The rating is calculated as a percentage of the Nebraska Criteria (Fischbach et al., 1982), which is listed as a footnote on Table D.3.
Table D.3. Pump-test data-summary percent of Nebraska criteria.
| Energy | No. | Ave (%) | Quartile average (%) | |||
|---|---|---|---|---|---|---|
| 1st | 2nd | 3rd | 4th | |||
| Natural Gas | 32 | 85.5 | 112.1 | 96.1 | 80.3 | 53.4 |
| Electric | 18 | 77.4 | 107.3 | 87.0 | 69.9 | 45.4 |
| Diesel | 17 | 69.9 | 97.8 | 81.2 | 66.4 | 34.2 |
| Propane | 4 | 47.3 | --- | --- | --- | --- |
| 71 total--Weighted average 77.3 | ||||||
|
Notes: 1. Nebraska Rating Criteria
|
2. Ten (10) tests were omitted from the summary. Six (6) were electric and four (4) were natural gas. Of the six (6) electric pumps, four were low yielding, <100 gpm; one was completed only in the High Plains aquifer; and one had defective data. Of the four (4) natural gas pumps: three were completed only in the High Plains aquifer and one had missing data. |
The criteria were developed after years of testing irrigation pumping plants to reflect the performance of a properly designed and maintained pumping plant. The criteria may be shown as an overall efficiency as in Table D.6, but for ease of use, are normally stated in units of fuel or energy input per unit of pump energy output, whp-hr (water horsepower-hour). Small differences in pumping-plant efficiency cause large differences in fuel or energy use and this method accurately reflects these changes by magnifying the differences in the ratings.
Table D.4 shows the average pump performance for each area and well type without regard to pumping-plant type. The distribution of pumping-plant types, however, does affect the results. All of the pumping plants in the southwest area were either electric or natural gas and all of the pumps for the southwest Dakota-High Plains subgroup were natural gas. As noted later, neither of these distributions would be considered typical for the area.
The distribution of pumping-plant types for the test group of pumps is in Table D.5 for each area and well type. The majority of the natural gas pumps are located in the southwest area; the majority of electrical pumps are in the north-central area and the majority of the diesel pumps are in the west-central area. The only groupings that were represented by pumps of all types were the north-central area and the group formed of all Dakota wells.
Table D.4. Pumping data summary by area.
| Area | Static level (ft) |
Dynamic Level (ft) |
Well Yield (gpm/ft) |
Nebraska Rating | |
|---|---|---|---|---|---|
| 1. Southwest | 240 | 277 | 774 | 85 | |
| (70-330) | (160-430) | (170-1230) | (40-120) | ||
| a. Dakota | 207 | 256 | 660 | 76 | |
| (70-260) | (160-430) | (170-900) | (50-105) | ||
| a. Dakota-High Plains | 246 | 280 | 858 | 89 | |
| (200-330) | (220-350) | (280-1230) | (40-120) | ||
| 2. West-central | 109 | 142 | 668 | 81 | |
| (30-220) | (40-280) | (400-1050) | (30-115) | ||
| 3. North-central | 49 | 98 | 432 | 61 | |
| (25-100) | (40-155) | (275-860) | (15-110) | ||
Table D.5. Distribution of pumping plants by energy source.
| Energy Source | A. Area | |||
|---|---|---|---|---|
| Southwest | West-central | North-central | Total | |
| Natural gas | 27 | 3 | 2 | 32 |
| Electric | 4 | 4 | 10 | 18 |
| Diesel | 0 | 13 | 4 | 17 |
| Propane | 0 | 0 | 4 | 4 |
| Total | 31 | 20 | 20 | 71 |
| Energy Source | B. Well Type | ||
|---|---|---|---|
| Dakota | Dakota- High Plains |
Total | |
| Natural gas | 9 | 23 | 32 |
| Electric | 18 | 0 | 18 |
| Diesel | 17 | 0 | 4 |
| Propane | 4 | 0 | 4 |
| Total | 48 | 23 | 71 |
Table D.6. Energy input per water horsepower-hour at 100% Nebraska rating criteria.
| Type | Efficiency (%) | Energy (kw/whp) |
|---|---|---|
| Natural Gas | 17.0 | 4.39 |
| Electric | 66.0 | 1.13 |
| Diesel | 22.7 | 3.29 |
| Propane | 19.1 | 3.90 |
Such a scheme makes internal combustion engines appear inefficient, but the efficiency of an electric motor is only part of the overall transformation of fuel into useful work. If a coal or natural gas burning power plant is 40% efficient and the transmission of electrical energy to the motor is 90% efficient, the overall efficiency of electrical pumping plants is only 23.8% which is comparable with other fuel-burning power sources used for pumping.
In order to determine the energy need for pumping where a large number and a variety of different types of pumping plants are used, both the distribution and condition of the various types of equipment must be known. This is usually done by sampling. A major difficulty, however, is that the required sample size to produce a given level of confidence in the data is seldom known. The size of sample needed to obtain a given degree of confidence depends on the variability of the parameter sought and its frequency of distribution. If the factor is highly variable and occurs infrequently in the population, a much larger sample is required to obtain a given level of confidence than if the factor is less variable or occurs more frequently. Unfortunately, neither the variability nor the frequency of occurrence is usually known prior to the study.
Sixty tests were requested for this study and a preliminary estimate indicated that 100 tests might be possible. Eighty-one tests were actually completed and 10 of these were deleted for various reasons. The sample size for this study was therefore fixed at 71.
Two types of data are included in the sample: continuous data which are represented by the pumping-plant ratings for which a mean and standard deviation may be computed; and proportional data which are represented by the distribution of the pumping-plant types. Cochran (1977) provides a method for estimating a sample size for each type of data (Appendix D.D). As might be expected, the sample size for each factor sought is different and increases rapidly with decreasing probability or increasing confidence limits.
The 71 tests of this study were used to examine the sample sizes needed to determine each of the various factors associated with the data (Appendix D.b). In general, the analysis indicated that the number of tests was adequate to describe the distribution of pumping-plant types at the 90% level but was too small for any greater level of confidence. The number of tests, however, was too small to adequately predict the pump ratings at the 90% level for all but the natural gas pumps and completely inadequate to describe the distribution of pumping-plant types in the individual study areas.
Thomas (1982) indicates that the distribution of irrigation pumping plants varies widely between counties and thus there is reason to believe that the distribution of pumping plants would vary between areas in this study. However, Thomas did not include public water-supply wells which are included in this study. To adjust for this, the distribution of well records for each county where tests were conducted was modified. The major portion (92%) was distributed as reported by Thomas (1982) and 8% of the total was added to the electrical pumps.
The basis for this adjustment was the proportion of public water supply wells relative to the combined total of irrigation and water supply wells in Kansas. Thomas (1982) reported that there were about 24,200 irrigation wells and Waldo (1989) reported that there were 2,100 public water-supply wells. Internal combustion engines normally are not used for pumping public water-supply wells; the motors are electrically driven. Unfortunately, no data on the distribution of public water-supply wells within the counties of the state were available which would have permitted further refinement of the estimates. The results are shown in Table D.7.
The effect of the various distributions is shown by comparing the energy inputs computed using a 100% Nebraska rating for all the areas and all 12 counties where tests were conducted with the energy input for the test-data distribution. As might be expected from the sample-size analysis, the energy value for all 12 counties is within 5% of the test data. The major difference is in the number of natural gas and diesel engines. The all-county data seem to be over represented by natural gas engines and under-represented by diesel engines.
The pump test data analysis for this study caused several concerns. In six years of pump testing the author has only encountered an occasional natural gas pump with a rating of 100%, and 11 of the 32 pumps tested for this study were above 100%. A second concern was that statewide only about one-third of the electrical pumps serve public water supplies but two-thirds of the electrical pumps (12 of 18) in this study serve public water supplies. A third concern was that past testing indicated that irrigation diesel-driven pumping plants normally have a higher rating as a group than natural gas pumps. The last concern is the poor showing of the propane pumps.
Table D.7. Comparison of pumping-plant type distributions
| Test Area | Gas (%) |
Electric (%) |
Diesel (%) |
LP (%) |
Energy Input (Kw) |
% of Test Energy |
|---|---|---|---|---|---|---|
| Southwest (6 counties) |
84 | 13 | 2 | 1 | 3.94 | 120 |
| West-central (3 counties) |
51 | 22 | 25 | 2 | 3.39 | 103 |
| North-central (3 counties) |
4 | 55 | 22 | 19 | 2.26 | 69 |
| All (12 counties) |
60 | 24 | 11 | 5 | 3.46 | 105 |
| Test Data (12 counties) |
45 | 25 | 24 | 6 | 3.28 | 100 |
The average performance data from Table D.3 are shown in Table D.8 with data from two studies of irrigation pumping plant tests on the High Plains of Texas (Agricultural Engineering Department, 1968; and Texas Department of Natural Resources, 1983). Both of these studies included more pumps than were tested in this study, but neither made an attempt to randomly select the equipment for testing. The earlier study attempted to select pumps geographically so not to leave any part of the study area barren of tests, but the later study did not describe site selection for tests. In addition, neither study included all four pump types.
The sample-size analysis in Table D.7. indicated that the performance rating for the natural gas pumps should have the highest level of confidence of any of the pump types tested, but the very high value (85.2%) appears to be suspect. The weighted average rating for the Texas data was only 68.1% by comparison or lower by 17.1%. A parametric comparison of the average values (Mendenhall and Ott, 1972) indicated that the difference was significant at the 95% level. The values, therefore, were drawn from two different populations and the Texas data do little to answer the concern. The electrical data, however, were verified. A parametric test of all three sets of electrical data indicated that the differences in the averages were not statistically significant at 95% level and the average is confirmed by additional data.
Table D.8. Comparison of pumping-plant performance (Nebraska rating criteria)
| Pumping Plant Type |
Values | Source of Data | Total Weighted Rating |
||
|---|---|---|---|---|---|
| Test Data | Ag Eng '68 | Texas '83 | |||
| Natural Gas | Number | 32 | 46 | 229 | 307 |
| Average Rating % | 85.5 | 63.3 | 69.1 | 69.9 | |
| Range % | (40.9-119.5) | (14.7-117.6) | (18.2-120.6) | (14.7-120.6) | |
| Electric | Number | 18 | 41 | 53 | 112 |
| Average Rating % | 77.4 | 73.6 | 79.7 | 77.1 | |
| Range % | (15.5-110.9) | (29.5-112.0) | (44.2-116.7) | (15.5-116.7) | |
| Diesel | Number | 17 | 26 | 43 | |
| Average Rating % | 69.9 | 88.3 | 81.0 | ||
| Range % | (20.8-108.0) | (61.7-110.0) | (20.8-110.0) | ||
| Propane | Number | 4 | 16 | 20 | |
| Average Rating % | 47.3 | 45.2 | 45.6 | ||
| Range % | (27.0-58.7) | (3.9-100.0) | (3.9-100.0) | ||
The additional diesel and propane data, however, were not analyzed. Both sets of data were too small to obtain good estimates of the standard deviation and, thus, further analysis was not performed. However, parametric comparison was made for the diesel data which indicated that the averages were statistically different. Finally, the propane pump data from the Texas studies had equally poor pump-performance ratings. The average rating in Texas was lower than the Kansas rating and the range of rating values was even greater. The only bright note was that at least one test resulted in a rating 100% for a propane-driven pumping plant.
The energy use for pumping is summarized in Table D.9. The distribution of pumping-plant types in the southwest area was used for both subgroups in this region. The energy-input values were computed using the average rating values from Table D.3 and are shown in two ways. The first is Kw which is really kwh/hr of operation and the second is watts/gpm-ft (the energy required to pump one gpm against a head of one foot). These values can be obtained by dividing the energy use rate (kw) by the product of the flow in gpm times the total head (lift + 2.31 x psi pressure) in feet.
The water-quality data were gathered as a service to others who are also working on the Dakota and Dakota-High Plains aquifers. It was also hoped that the analysis of the inorganic constituents might prove useful in identifying the source aquifer. Unfortunately, the data shown in Appendix D.C are of very limited usefulness for this study. Wells in close proximity show similarities in water quality but are quite dissimilar to waters from the same aquifer at a distance.
Table D.9. Pumping-energy need or use
| Area | Flow Q (gpm) |
Pump lift (ft) |
Outlet press (psi) |
Pump output (whp) |
Energy use rate | |
|---|---|---|---|---|---|---|
| Kw | watts/gpm-ft | |||||
| 1. Southwest | 741 | 280 | 26 | 63.6 | 298 | 1.18 |
| a. Dakota | 571 | 260 | 32 | 48.1 | 225 | 1.18 |
| b. Dak-H.P. | 798 | 287 | 24 | 60.0 | 323 | 1.18 |
| 2. West Central | 670 | 153 | 37 | 40.3 | 173 | 1.08 |
| 3. North-central | 462 | 99 | 39 | 22.1 | 80 | 0.91 |
Missing values for measurements of temperature were primarily due to oversights during the pump-testing program but most of the missing pH data were due to a meter failure during the course of the summer.
Another problem was related to the electrical power measurements. A wide variety of meters are installed by the various power companies and some do not allow for field power measurements by timing the drive disc revolutions. A now-available portable watt-hour meter that will measure wattage, volts, amperage and power factors would solve this problem.
A change in procedure is also needed for recording static water levels. Normally, the static water level is measured prior to starting the test but not always. The difficulty occurs when the specific capacity is low. Recovery of the true static water level may require hours or days instead of minutes and, if pumping has occurred within the necessary recovery time, the measurement can be in serious error. The solution is to make the static water-level measurement prior to any pumping, inquire when the pump was last operated and make a notation on the test record sheet of the circumstances found. Many of the specific-capacity calculations made for these tests seemed erroneous.
The results of pumping-plant performance tests vary with time and need to be performed several times during the season as pumping lowers water levels in the aquifer. These tests should have been made during the active part of the irrigation season. However, unusual rains in the southwest resulted in the tests being run at the start-up of the irrigation season for many and early to mid season for the remainder. An attempt to determine how many hours the pumps had been run prior to the test should have been included in the field data, but was not. The tests, however, were performed for the most part with few mistakes or omissions in the data gathered.
The evaluation of energy needs was the most interesting and, yet, most troublesome aspect of this study. Past work has assumed or calculated an average rating or efficiency for pumping without exploring the interaction of pumping-plant type distribution and ratings. Unfortunately, the distribution of pumping-plant types was not considered originally. Not until the pump ratings were combined with several distributions did the necessity of having a good estimate of the proportions of the different types of pumping plants become apparent. Fortunately, some data were available in the literature from which estimates could be made. The procedure used, however, was not exact. The test distribution according to the sample-size estimate should have been very close to the distribution for all 12 counties (Table D.9). The effect was not great when combined with a 100% Nebraska rating but as noted earlier seemed to have too many natural gas pumps and too few diesel pumps for a sample that was 17% of the population.
Improvement in the energy use values undoubtedly could be made with more pump testing and more data on pumping-plant types. A random survey by phone could answer the question of pump type distribution without having to return to the field. The owners undoubtedly know the type of pumping plant used and would most likely know whether the discharge is greater or less than 100 gpm. A careful screening of the well records, however, prior to any additional study would be helpful in eliminating wells that are neither Dakota nor Dakota-High Plains.