Improving Uniformity of Overhead Irrigation Systems to Reduce Water Use and Maximize the Retention of Nutrients in Container Grown Nursery Crops

Transcription

1 Improving Uniformity of Overhead Irrigation Systems to Reduce Water Use and Maximize the Retention of Nutrients in Container Grown Nursery Crops Water Adaptation Management and Quality Initiative January 2015 Prepared for: Farm & Food Care Ontario 100 Stone Road West, Suite 106 Guelph, ON N1G 5L3 Prepared by: Dr. Jeanine West PhytoServ 6 William Drive Cookstown, ON, L0L 1L W A M Q I 4 1

2 Executive Summary Efficient use of irrigation water in horticultural production systems is a high priority for research in Ontario horticultural crops. The nursery sector is looking for ways to reduce total water applied while improving irrigation application uniformity. Overhead sprinkler systems are the most common form of irrigation for container nursery crops (#1-5 pot size) because overhead systems are relatively inexpensive to install, requires minimal maintenance (i.e. less labour) and can be used to cool the plants in the heat of summer. However, in outdoor applications, overhead sprinklers often produce patterns of uneven water application that lead to inconsistent water and nutrient uptake - affecting the quality and consistency of product. This research evaluated nozzle types, operating pressure and irrigation layouts (central bed design vs. peripheral design) in outdoor container growing systems in order to improve irrigation delivery uniformity across the zones and reduce the total water applied. Considering that low operating pressures during the test runs may have influenced the results at one of the test sites, the data indicates that Site B s modified peripheral design (peripheral line of brass head sprinklers at edge of entire block with single line of traditional brass sprinklers per bed, offset and in an alternating pattern with no driveways between beds) at high operating pressure was the best irrigation design.. The extra peripheral sprinklers resulted in plants receiving water from more than two sprinklers, which increased the interception significantly and allowed for a shorter irrigation time. The intentional staggering and overlap in sprinkler patterns seemed to maximize the overall distribution uniformity. Further tests at Site A with plastic sprinklers at higher pressures and with a greater overlap area should be performed next season to better evaluate the efficiency of the Nelson sprinkler heads. Purpose The purpose of this study was to assess sprinkler pattern layout (traditional central bed design vs. new peripheral design) and pressure to increase uniformity and decrease the application period length. The expected result would be a reduction in total water applied and reduced incidence of nutrient losses through leaching. The specific objectives of this study were to maximize: Efficient water use to minimize the operation s demand on the water resource Distribution uniformity across the nozzle application area 2 W A M Q I 4 1

3 The retention of nutrients in the container to improve nutrient use efficiency, production uniformity and reduce impacts of runoff water quality Methods Demonstration Sites Most traditional overhead irrigation layouts are of a central bed design, consisting of foot long beds that are 18 feet wide with 14-foot driveways on either side. Usually, the irrigation sprinkler risers have 360 o pattern heads placed 30 feet apart along the centre of the bed. Sprinkler heads on the ends are often replaced with 180 o heads. There are no sprinklers on the other two sides of the block (known as peripheral design ). This sprinkler layout results in dry zones at the corners and edges of the beds due to the radial sprinkler pattern and factors like wind. Some plants are receiving irrigation from only one sprinkler, while others are receiving irrigation from two sprinklers. Containers on the windward side of the bed (usually southwest) receive even less water when winds are greater than 2.2m/s. Site A is a container nursery farm of approximately ten hectares, growing a mix of evergreens and shrubs, with a traditional overhead sprinkler layout of central bed design for their coldframes (also called hoop houses or poly houses). The coldframes are 18 wide, with 15 on each side of growing area, and another 12 for a driveway before the adjacent coldframe (see Figure 1 top left). The coldframes average 300 long. Irrigation sprinklers are placed on risers (posts) laid out in one row down the centre of the bed (e.g. Figure 1 (top left), Figure 2a Bed A14). Risers are spaced 30 apart, fed by a 1.5 supply line for the first 100, switching to 1 for the remaining 200. The standard sprinkler for the growing area is the full circle impact brass Rainbird 20JH model, fitted with a 3/32 nozzle (slightly smaller orifice than typical to decrease overall water volume applied). A 4 header line feeds the irrigation system from the main pump. The nursery farm waters on a zone basis with coldframes watered at once to maintain a minimum design pressure of 35 psi. At Site A, 3 different peripheral designs were constructed to compare to the traditional bed design. Different sprinkler nozzles were laid out at the edge of the bed, spaced 20 and 30 feet apart. These new peripheral designs were then compared to the traditional bed design and evaluated for distribution uniformity and pressure. 3 W A M Q I 4 1

4 Site B is a separate container nursery farm of approximately ten hectares, growing a mix of evergreens and shrubs, with a slightly modified sprinkler layout that is a modified (offset risers) peripheral design of sprinklers. While the coldframes are 18 wide, the space between coldframes is only 12 and the space is entirely used for planting (no driveways). Site B s coldframes were 572 long. The typical sprinkler layout at this farm is in one row per bed, however the spacing between risers within row is 40. The risers between rows (adjacent coldframes) are 30 apart, with the risers offset, in a triangle pattern (Figure 1 top right, Figure 3). The central bed sprinklers are traditional full impact brass Rainbird 30H sprinklers, fitted with a standard 5/32 nozzle (red dots, Figure 1 top right). Part-circle (180 0 ) sprinklers (Rainbird PJ with 5/32 nozzles) are used at the ends of each coldframe and at the periphery of the entire growing block (semi-transparent blue areas around each riser, Figure 1 top right). A 6 header line feeds the irrigation system from the main pump. The nursery farm waters on a zone basis with approximately 8 coldframes watered at one time to maintain a minimum design pressure of psi. Factors affecting efficiency and uniformity Design factors: Sprinkler type At Site A, Rainbird 20JH traditional brass sprinklers were compared with Nelson R10TJ plastic rotator heads fitted with both black (5 spray streams) and green (1 spray stream) plates. Site B was a benchmark test site, with only Rainbird 30H sprinklers used. Nozzle type - At Site A, 3/32 nozzles were used on the Rainbird brass sprinklers, and the red 1/8 nozzles were used on the Nelson rotators. Site B was a benchmark test site and an above-average industry comparison to Site A, with 5/32 nozzles used on all heads. Sprinkler spacing (includes compact/traditional, linear/alternate) - At Site A, the Rainbird and Nelson rotators were initially tested all at 30 spacing down each row (Figure 1 bottom right, Figure 2a centre). In August and September, the design was changed to test the Nelson rotators with both plates at 20 spacing (Figure 1 bottom centre-right, Figure 2b). Site B was a benchmark test site, with their unique alternating layout of 30 4 W A M Q I 4 1

5 between rows and 40 between risers in one row (Figure 1 top right, Figure 3). Sprinkler Layout In addition to comparing Site A s (traditional) central bed design with linear (opposite) pattern to Site B s peripheral design with alternating (triangle) pattern, Site A s central bed design was compared to two rows peripheral rows along the outside of the coldframes (Figure 1 left and centre-left, Figure 2a left versus right). Operational factors: Pressure Site A s operating pressure ranged from psi during the trial. Late into the trial it was discovered that the typical operating pressure in a normal irrigation event was approximately 35 psi. The number of growing beds irrigated at one time influences the overall pressure available. Site B s operating pressure was typically around 55 psi, and was not altered during the test period. Length of irrigation cycle The irrigation cycles at both sites depended on the crops need for water. While Site A may irrigate some crops every day and even twice a day (especially established plants), newly potted plants may only receive water every other day. The plants are usually watered over a 2-3 hour period, depending on the need. For the purposes of this study, the tests were run at 30-minute intervals. At Site B, similar concepts apply, but irrigation events generally occurred in 12 to 15-minute increments (cyclic/pulse watering). Site B s irrigation system is fully automated, allowing them to achieve precise cyclic/pulse watering without additional labour. At Site A, the first tests were conducted at 30, 60 and 137 minute intervals, and at Site B, the tests were conducted at 34 and 60-minute intervals. Wind and other environmental conditions (> 2m/s) Wind speeds on all test days varied greatly, from less than 0.5 m/s to over 7 m/s gusts. Only maximum wind speeds are reported; however, the measurements of wind speed & direction, temperature, and barometric pressure were recorded hourly through the testing period. 5 W A M Q I 4 1

6 Crop Spacing -Note that crop spacing was not investigated, as the study areas contained mature #2 or #3 potted material, fully spaced. The crop spacing will affect interception efficiency, not the irrigation uniformity. Measures of Uniformity, Leaching Fraction, Flow and Pressure Uniformity tests were run at Site A on four dates (July 17 & 30, August 20, and September 23, 2014). Site B had two test days on August 14 th and November 3 rd, Catch can tests were used to calculate the Distribution Uniformity. Both the Lowest Quarter DU (DUlq) and Christiansen s Uniformity Coefficient (CUC) were calculated, as well as determining the Nomograph ranking. Pots were laid out in a 5-foot grid pattern, radiating out from a central sprinkler head (see blue boxes in Figures 1-3 for test areas). At least 20 pots were used for each test, with the volumes listed in increasing order before removing the lowest 25% volumes (following the protocol of Dudek and Fernandez (Michigan State). Areas that contained plants with large canopies were avoided, and plants near sprinkler heads and catch cans were moved to avoid canopy interception of water. All catch cans were leveled to account for bedside slope. The pattern of high and low volumes generally followed the same pattern across the repeat runs. Leachate fractions were tested by setting one empty pot beside a test plant, both pots contained a plastic liner to capture both irrigation volume and leachate volume, respectively. The test plant was placed on a block inside the capture pot, to ensure that any leachate from the irrigation event would not be re-absorbed into the plant. On the August test date, leachate was collected and sent to A&L Laboratories (London, Ontario) for full ICP-MS analysis. The leachate fractions were compared to weight changes in the pots (weighed before and after irrigation). Flow was determined by measuring the amount of time (in seconds) to capture 4L of water coming from each sprinkler head with a tube. The capturing pail was marked at the 4L level, and holes were drilled at the height to increase the visibility of these measurements during the trial. Pressure was determined with a pitot tube (Vanden Bussche Irrigation), fitted with either a 0-100psi or 0-60psi oil-filled pressure gauge (depending on operating pressure on that test day). The tip of the tube was inserted into the nozzle aperture, and the reading was taken when the pitot tube blocked all of the nozzle flow. Where applicable, 6 W A M Q I 4 1

7 a foam block was placed around the tube to prevent the tip from entering too far into the nozzle and disrupting the plastic inserts. One site had removed the inserts because of plugging issues while the other had theirs in place (municipal water source). Because of the close fit, it was particularly difficult to check the pressure of the Nelson rotator heads with the pitot tube. Results & Discussion Distribution Uniformity In 2013, OMAFRA and AAFC researchers were able to demonstrate the distribution uniformity of overhead impact sprinklers in the field to be about 40-50% (PhytoServ 2014a). Industry standards cite 60% as the lowest acceptable threshold for distribution uniformity, with 75% the upper limit of efficiency for this form of irrigation equipment. Catch can tests were used as a tool to determine distribution uniformity at the two sites studied in this project. The results of the catch can tests are summarized in Table 1, showing a range of % uniformity (DUlq) depending on the layout, sprinklers/nozzles, and length of test. Christensen s Uniformity Coefficient (CUC) was calculated to be the same as the lowest quarter Distribution Uniformity (DUlq). The Nomograph Ranking was a third way to categorize the uniformities from the catch can tests, and the results generally matched the DUlq results. The best DUlq was observed at Site B with the peripheral design and triangle riser pattern on August 14 th (74.7%), but repeats of this study on November 3 rd resulted in DUlq only slightly above (66.4%) the average for all test layouts (59.1%), due to high winds. The worst performance (55.2% DUlq) was observed at Site A from the peripheral design and opposite riser pattern with Nelson sprinklers fitted with black plates at the 30 spacing, but the single line Rainbird sprinklers did not perform significantly better (55.8%). In fact, the Nelson sprinklers fitted with the green plates had the next best performance compared to the Site B layout with DUlq averaging 60.9%. The pattern of volumes across the layout of cans in each of the test layouts at Site A (Figure 4) illustrate visually the location of pots across a growing bed that received the least (blue) and most (orange) amount of water (lowest quarter and highest quarter shaded). The wind impact can be seen in the fourth row of boxes (July 30 data): as the 7 W A M Q I 4 1

8 maximum wind speed increased to 2.1m/s, the catch cans with the lowest volumes (blue shaded numbers) are all on the south side, where the impact of the eastward (and slightly northward) wind would have more influence. Comparing the initial three layouts at Site A (as in Figure 2a), the highest amount of water appears to be applied down the centre of each bed, with some skewing to the north side of the beds, although the pattern is not consistent. When comparing the Nelson sprinklers (with different plates) at the 20 spacing (Figure 4, bottom 4 blocks), there is no clear pattern, however, the wind effect seems to be less pronounced. Lowest volumes were evident on the south side and in the centre of the beds on August 20 th when the maximum wind speeds exceeded 2 m/s, while the September 23 rd results when there was little wind (0.8 m/s max speed) have lower volumes more evenly distributed across the centre of the growing area. At Site B, strong west winds (over 7 m/s) did not seem to impact the location of the lowest volumes in the catch cans (Figure 5). While there is some evidence of higher volumes along the centre of the growing bed, the pattern is not consistent between the different test runs. The peripheral layout, triangular nozzle pattern, higher output and pressure of these sprinklers/nozzles likely compensated for the uneven sprinkler pattern and the higher uniformity for Site B. Leaching Fraction and Nutrient Analysis Leaching Fraction percentages at Site A ranged from % (Figure 6), far greater than typical leaching fractions of 5-30% expected. The frequent and extensive rain events during the 2014 growing season meant that some testing was carried out at Site A when crops were already saturated. With sub-optimal distribution uniformity and variations in canopy it is possible that the empty pots did not receive the same amount of irrigation water as their neighbouring plants, although the pots were placed adjacent to avoid this variable. Nutrient analysis was conducted on pooled samples of leachate (n=9), on-farm drain and recycling pond water at Site A (Table 2). Nutrient concentrations (ppm) were then compared to MOE Storm Water Guidelines (not shown). Nutrients of historical concern by the MOE include nitrate-nitrogen, total phosphorus as well as metals. At Site A, all three water samples came back at 0 ppm for nitrate-nitrogen, well under the MOE threshold 10 ppm. For total phosphorus, all three water samples came back well 8 W A M Q I 4 1

9 under the MOE threshold of 0.5 ppm. All other nutrients were less than the MOE guidelines and based on data from other research, the leachate at Site B (data not shown) also poses a very low risk to the environment (PhytoServ 2014b). Leaching Fraction percentages at Site B ranged from 0-48% with an average of 5% (Table 3). A Leaching Fraction of <10% is considered to be quite low. Site B conducted very conservative irrigations (2 x 17 minutes) based on ET models and the pulsing of the irrigation events resulted in much more efficient wetting of the root zone. Because of the irrigation BMP s in place at this nursery, a very low Leaching Fraction was achieved. The length of time between irrigation events (cyclic irrigation) would allow more time for plants to take up the water, also ultimately decreasing Leaching Fraction at this site. Pot weights were also carried out at Site B to see if the water added through the irrigation even could be quantified by weight and used as a tool by the grower to fine tune irrigation cycle length and timing for the crop. Weights were recorded before and after irrigation on a variety of plants. The difference in these plant weights represented the amount of water that the media absorbed in grams. Each gram difference represents 1ml of water. Through this project, we were able to demonstrate to the grower that, in addition to traditional crop monitoring, difference in weight (before and after irrigation) can be an excellent tool in measuring irrigation effectiveness and uniformity throughout the bed, identifying exact locations of excess or insufficient water. Container wetting front, media and canopy inspection Crops were inspected at the end of each irrigation period. The grower noticed over-application in the double brass impacts, and the donut effect of watering with the Nelson double rows (especially with the black plates) where too much water was applied at the nozzle and furthest from the nozzle with less in the middle. Referring back to the sprinkler patterns illustrated in Figure 1, the observation that the double rows of sprinklers (especially brass impacts) covered a lot of non-growing areas at Site A, including laneways and nearly over to the next growing area. At Site B, while there are areas with 4 overlaying patterns, generally the overlap areas are quite small relative to the overall spray area, and the elimination of laneways increased the effective interception area substantially. 9 W A M Q I 4 1

10 Observed radii of the sprinkler/nozzle combinations very closely matched the manufacturer s ratings (see Table 1), although it is important to note that distinct pattern changes were observed in all Site A sprinkler patterns when the winds exceeded 2 m/s. Increasing droplet size, volume applied, and using the typical lowangle upright spray pattern (as opposed to the high angle/multi pattern provided by the black plates of the Nelson sprinklers) appeared to be the best way to resist wind effects. Pressure at sprinklers, output volumes At Site A, testing demonstrated a loss in pressure (psi) as the distance from the header increased (Figures 7 & 8). The difference in pressure was very small (4-14%) and did not result in significant changes to nozzle output and distribution uniformity. Conservation of sprinkler pressure was probably due in part to a reduction in the irrigation pipe used, part way down the line (see Figure 2a right), as illustrated by the correlation graphs between flow and pressure (Figures 7& 8 top) for each layout. On both test dates, the two rows of brass Rainbird sprinklers correlated in a negative manner (decreasing flow with increased pressure), while the single row Rainbird and Nelson layouts had essentially flat correlations. Figure 9 represents the correlation between flow and pressure for the Nelson sprinklers (both with green and black plates), with similar results to previous test dates. Preliminary tests at other nurseries (PhytoServ 2014a) suggest that most container beds lose significant amounts of pressure as measured further away from the header. At Site B, testing demonstrated a loss in pressure (psi) as the distance from the header increased (Figure 10 bottom). The difference in pressure was very small (8-9%) in the first quarter of the bed and did not result in significant changes to nozzle output and distribution uniformity. This confirms preliminary tests at other nurseries suggest that most container beds lose significant amounts of pressure as measured further away from the header, although there was more variability in the flows with higher pressures (Figure 10 top). At Site B, across the length of the entire 580-foot bed, pressure dropped 20-30% on both test dates. Interestingly, the output (L/min) only dropped 11%. The drop in psi at the far end of the bed resulted in minor decreases in nozzle output and slight decrease in distribution uniformity, likely due to the significant nozzle size and sprinkler head design. 10 W A M Q I 4 1

11 The total water applied per area was also determined based on output volumes of the sprinklers (see Table 1). Typically, much more water (20-30 mm/event) is applied to compensate for the dry zones in the irrigation sprinkler pattern, but researchers have found that 15 mm/event (Danelon et al. 2010) should be adequate for water absorption in container media, a number corroborated by nursery growers. As is evident in Table 1, the high pressure, triangle sprinkler layout and peripheral design at Site B provided the most water per unit area (over 1700 L/ha/min), consistent with our findings and the manufacturer s ratings. Interestingly, the Nelson R10TG s fitted with the green plates provided a large volume of water to the growing area as well, with approximately 1600 L/ha/min. Grower observations After looking at the data, it was determined that initial test pressures at Site A were inadequate for thorough evaluation of optimal performance of the various sprinkler types and layouts. At Site A, Coldframe #A12 was set up all season with the Nelson R10TJ heads on black plates (30 spacing), and higher pressures did result in better distribution uniformity (Table 1). After the trial was complete, the farmer observed that at even higher pressure the nozzles gave improved distribution of water over the crop. In fact, the Spirea Little Princess crop had the best consistency ever achieved at this nursery, under this new irrigation setup, noted by several employees at the end of the season. Another aspect that may have affected the study results was the amount of water applied and required by the crop. In general, the crops used in the test beds (A10, A12, A14) were large, all in their second year of growth, and were fully rooted throughout the pot, making them more difficult to irrigate with overhead sprinklers. The studies were also hampered by the frequency of heavy rainfalls experienced through the summer leading to very wet crops and subsequently negligible wetting/irrigation event impacts on the test days. At Site A, the farmer reported that they have learned a lot during the course of this research. The farmer will be adjusting their watering patterns and irrigation setup based on the results of this study. They learned the importance of using adequate pressure, and the loss of pressure down the length of the bed. The farmer also intends 11 W A M Q I 4 1

12 to replace their water lines with 2 piping throughout (replacing the lines) to allow for increase water volumes to reach the end of the coldframe. This change, combined with the increased pressure will create the best scenario for a shorter time frame of watering with the best distribution uniformity. From the tools and procedures they learned through this project, the farmer plans to continue to evaluate their overhead irrigation system for uniformity, effectiveness and efficiency in the years to come. They will continue to evaluate the uniformity of traditional single-line brass nozzles and the new layout Nelson R10TJ heads with green plates, at the 20 spacing which they are adopting for use at their other farm. Site B is satisfied with the results of the study at their farm as it confirms their earlier self-audits about sprinkler performance. Site B spent several years evaluating their overhead irrigation systems and making adjustments (e.g. adding an extra line of 180 o nozzles on the upwind side) to increase uniformity and effectiveness. The results in this study confirmed that Site B is running at settings that allow for above average performance of traditional single-line brass sprinklers. General recommendations based on the results include 1. DU calculation may not completely explain all of the parameters that affect how evenly and effectively plants are being watered 2. Weather conditions (e.g. wind) has a major impact on both DU studies and irrigation management 3. Several other factors can be used to interpret the efficiency of an irrigation system 4. Careful crop monitoring (e.g. wet/dry spots) were not included in this project, but could have added more insight into the research. 5. Each system must be fine-tuned independently because of differences in a variety of parameters (e.g. crop architecture, pressure, number of beds to be watered, difference in irrigation style and media etc.). 12 W A M Q I 4 1

13 Overall table of results (combined farmer/researchers): Sprinkler/Layout Pros Cons Rank Brass 20JH, 1 line, oppositely spaced Brass 30H, 1 line, alternately spaced Nelson R10TJ Green Plate, 2 lines, oppositely space Brass 20JH, 2 lines, oppositely spaced Nelson R10TJ Black Plate, 2 lines, oppositely spaced Best overall performance if coldframes have driveway between Good volume Better DU at higher pressure Greater volume Better DU at higher pressure If no driveways, best performance More consistent watering pattern than the black plate Minimal water on the side driveways Greater volume Better DU at higher pressure Unique spray pattern (5 streams, one long, rest varied) designed to increase consistency across spray radius Spirea crop watered with this design for the entire season had the best consistency compared to all historical crops Ends and edges require spot watering or extra peripheral line of 180 o nozzles Ends and edges still require some spot watering or extra 2 peripheral line of 180 o nozzles Evidence of overapplication furthest from the nozzles 3 Spray reaches too far into laneways/next bed at required high pressure, not designed for this pattern Insufficient water at the pressure tested Evidence of overapplication furthest from the nozzle Alternative strategies Some growers (e.g. Site B) will spot-apply supplemental water to the dry zones to compensate for the reduced irrigation interception in those bed locations. Practices such as spot watering will result in substantial savings in water consumption, less contaminated runoff water and more consistent fertilizer uptake. This can be achieved by watering dry zones with a mobile boom and or hand wand. Spot watering is often too labour-intensive and therefore not a common practice. Of course, using micro- 13 W A M Q I 4 1

14 irrigation (e.g. drip, spray stakes), capillary mats or ebb and flow systems will result in more even water application but these systems are often not economically or physically appropriate in outdoor container production. According to irrigation equipment suppliers, Roberts provides a 3-way sprinkler that may have better distribution uniformity in the field, although the primary school of thought is that closer spacing is the key, and new irrigation layouts are being designed with greater overlap zones. Communication of Results or KTT An article was prepared for the Landscape Ontario HortTrades magazine, the final report will be posted online on the Landscape Ontario website (growers page), and the research will be presented on February 4, 2015 at the Ontario Nursery Growers Short Course. The research in this study will also be shared through individual communications as part of the Wastewater Strategy Project funded by OFIP to Landscape Ontario. Acknowledgements The research team would like to extend their appreciation for the funding of this project through the Water Adaptation Management and Quality Initiative, administered by Farm and Food Care Ontario. The project s success is due to the support of the Grower s Group of Landscape Ontario and the farmer co-operators across Ontario that participated in this study. The research team (Jennifer Llewellyn OMAFRA, Wade Morrison, Shannon Gauthier) were invaluable in supporting the experimental design, data collection and analysis. References Danelon M, A Kachenko, J McDonald, C Rolfe & B Yiasoumi Nursery Industry Water Management Best Practices Guidelines. Nursery & Garden Industry Australia. Dudek and Fernandez. Conducting a water application uniformity evaluation for an overhead sprinkler irrigation system in the nursery. Michigan State University Extension (no date). PhytoServ 2014a. Water Balance Case Study at an Outdoor Ornamental Nursery. Farm & Food Care WRAMI # 17. PhytoServ 2014b. Outdoor Container Nursery Production Water Use Efficiency and Best Practices Benchmarking Study. Farm & Food Care WRAMI # W A M Q I 4 1

15 Figure 1 Standard central bed design for Site A (left) compared to the modified peripheral design of site B (right, only partially illustrated). Red dots = sprinklers, Blue dots are sprinklers. Site A: Standard Layout (left), and test layouts: Brass 2 row layout (centre left), Nelson green 20 (centre right), Nelson black 30 (right) 15 W A M Q I 4 1

16 Figure 2a Sprinkler/Bed Layout for Site A (first tests) Site A layout: New Layout with 2 rows brass sprinklers (red) Bed A10, New Layout with Nelsons in 2 rows (grey/black) Bed A12, Traditional Layout with 1 row brass sprinklers (red) A14. Not to scale. Blue Box indicates test area for distribution uniformity. Site A Layout 1 Cold Frames ` Lane Way Lane Way N A14 A12 A10 4 Header 16 W A M Q I 4 1

17 Figure 2b - Sprinkler/Bed Layout for Site A (second tests) Site A layout 2: Condensed layout Aug 20/Sep 23 in Bed A10, 20 spacing with first 4 sprinklers Nelson R10TG with green plates (green/black), remaining sprinklers Nelson R10TG with black plates (grey/black). Not to scale. Site A Layout 2 Cold Frames ` Lane Way Lane Way 30 N A12 A10 A8 17 W A M Q I 4 1

18 Figure 3 - Sprinkler/Bed Layout for Site B Layout at Site B West Farm Site B Risers Risers Risers ` 12` ` 8` 8` 18` 18` 18` Bed 39 Bed 38 Bed W A M Q I 4 1

19 Table 1* - Summary of Distribution Uniformity and Parameters Impacting DU for Both Site A and Site B Farm Date Sprinkler head Nozzle # lines Bed design - layout Sprinkler pattern Length of bed (ft) Spacing (ft) Max wind speed (m/s) Pressure (psi) Rated US GPM Rated Radius (ft) Observed Radius (ft) Length of run (min) Total water applied (L/ha/min) DU(lq) CUC Site A 17-Jul Nelson R10TG plastic with black plate red 2 west-east opposite % 54.2% Poor Site A 17-Jul Nelson R10TG plastic with black plate red 2 west-east opposite % 54.5% Unacceptable Site A 17-Jul Nelson R10TG plastic with black plate red 2 west-east opposite % 55.2% Poor Site A 30-Jul Nelson R10TG plastic with black plate red 2 west-east opposite % 68.6% Unacceptable Nomograph Ranking Average DUlq 55.2% Comments Site A 20-Aug Nelson R10TG plastic with black plate red 2 west-east opposite % 52.1% Unacceptable 2nd run only 57.2% Site A 23-Sep Nelson R10TG plastic with black plate red 2 west-east opposite % 62.3% Fair Site A 20-Aug Nelson R10TG plastic with green plate red 2 west-east opposite % 60.5% Fair 2nd run only 60.9% Site A 23-Sep Nelson R10TG plastic with green plate red 2 west-east opposite % 61.2% Poor Site A 17-Jul Rainbird 20JH brass impact 3/32" 1 west-east opposite % 62.9% Poor Site A 17-Jul Rainbird 20JH brass impact 3/32" 1 west-east opposite % 56.4% Poor 55.8% Site A 17-Jul Rainbird 20JH brass impact 3/32" 1 west-east opposite % 35.5% Unacceptable Site A 30-Jul Rainbird 20JH brass impact 3/32" 1 west-east opposite % 68.6% * inverted & lower times? Site A 17-Jul Rainbird 20JH brass impact 3/32" 2 west-east opposite % 50.7% Unacceptable Site A 17-Jul Rainbird 20JH brass impact 3/32" 2 west-east opposite % 61.6% Poor Site A 17-Jul Rainbird 20JH brass impact 3/32" 2 west-east opposite % 52.0% Poor Site A 30-Jul Rainbird 20JH brass impact 3/32" 2 west-east opposite % 71.3% Fair 58.9% Site B 14-Aug Rainbird 30H brass impact (ends are PJ) 5/32" 1 north-south triangle/alternating % 74.7% Fair Site B 03-Nov Rainbird 30H brass impact (ends are PJ) 5/32" 1 north-south triangle/alternating % 56.3% Poor 66.4% Site B 03-Nov Rainbird 30H brass impact (ends are PJ) 5/32" 1 north-south triangle/alternating % 63.8% Fair Site B 03-Nov Rainbird 30H brass impact (ends are PJ) 5/32" 1 north-south triangle/alternating % 70.9% Fair central main overall = 59.10% * This page prints on legal size paper 19 W A M Q I 4 1

20 Figure 4 Site A Distribution Uniformity Site A Catch Can Test Volumes (ml) for Distribution Uniformity. Blue highlighted values have the lowest volumes, orange highlighted values have the highest volumes, and the red highlighted volume was determined to be an outlier. The solid black lines running horizontally through the blocks illustrate the sprinkler rows. North A14 (1 row brass) A12 (2 rows plastic/black plates) A10 (2 rows brass) July 17/30min Max Wind=1.5m/s Pressure=23.5psi DU(lq)= 62.9% DU(lq)= 54.1% DU(lq)= 50.7% July 17/60min Max Wind=1.5m/s Pressure=23.5psi DU(lq)= 56.4% DU(lq)= 54.6% DU(lq)= 61.6% July 17/137min Max Wind=1.5m/s Pressure=23.5psi DU(lq)= 35.5% DU(lq)= 55.2% DU(lq)= 52.1% July 30/30min Max Wind=2.1m/s Pressure=30psi DU(lq)= 68.6% DU(lq)= 56.8% DU(lq)= 71.2% 20' Spacing, Green Plates 20' Spacing, Black Plates (designed for 30') Aug 20/36min Max Wind=2.3m/s Pressure=24psi DU(lq)= 60.5% DU(lq)= 52.1% Sept 23/60min Max Wind=0.8m/s Pressure=37psi DU(lq)= 61.2% DU(lq)= 62.3% 20 W A M Q I 4 1

21 Figure 5 Site B Distribution Uniformity Site B Catch Can Test Volumes (ml) for Distribution Uniformity. Blue highlighted values have the lowest volumes and orange highlighted values have the highest volumes. The solid black lines running horizontally through the blocks illustrate the sprinkler line down each growing bed. West Site B - West Farm (irrigation main on S end) S end of bed, about 1/3 along about 2/3 along bed Aug 14/34 min Max Wind=7.3m/s Pressure=47-59psi DU(lq)= 74.7% Nov 3/34min Max Wind=6m/s Pressure=48-56psi DU(lq)= 56.3% DU(lq)= 63.8% Site B - East Farm Just S of the central irrigation main Aug 14/60 min Max Wind=7.2m/s Pressure=64-67psi DU(lq)= 70.9% 21 W A M Q I 4 1

22 Figure 6 Site A Leachate Fractions 22 W A M Q I 4 1

23 Table 2 Site A Leachate Nutrient Content n=1 n=1 n=9 Parameter Units Pond On-Farm Drain Average Leachate StdDev Leachate Adjusted SAR Nitrate-N ug/ml Sulphur (as SO4) ug/ml Aluminum ug/ml Boron ug/ml Calcium ug/ml Copper ug/ml Iron ug/ml Potassium ug/ml Magnesium ug/ml Manganese ug/ml Molybdenum ug/ml Sodium ug/ml Phosphorus ug/ml Silicon ug/ml Chloride ug/ml Ammonia (NH3/NH4-N) ug/g phc Residual Sodium Carbonate Saturation Index Total Alkalinity ug/ml Anion Sum Meq/L Bicarbonate ug/ml Carbonate ug/ml Cation Sum Meq/L Conductivity (@ 25 deg C) ms/cm Hardness ug/ml ph SAR Total Dissolved Solids ug/ml Phosphorus (H2PO4) ug/ml Zinc ug/ml *green-highlighted cells have numbers that are BDL (assumed zero) 23 W A M Q I 4 1

24 Table 3 - Site B Weights versus Traditional Leachate Test Before Weight (Grams) After Weight (grams) Difference Leachate Empty Pot # (grams) (ml) Pot (ml) % Location % Middle % West % West % Middle % Middle % East % Middle % Middle % West % Middle % East % East % Middle % East % East % Middle % West % West % West % east % east % Middle % west % Middle average: W A M Q I 4 1

25 Figure 7 Site A Flow and Pressure (July 17, 2014) 0.4 Bed 10 Flow and Pressure 0.4 Bed 12 Flow and Pressure 0.4 Bed 14 Flow and Pressure Flow (L/s) 0.2 Flow (L/s) 0.2 Flow (L/s) Pressure (psi) Pressure (psi) Pressure (psi) A Beds July 17, 2014 Pond Lane Way A-10 A-12 A Pump House North Legend Lateral Line (psi) (L/min) (psi) (L/min) (psi) (L/min) (psi) (L/min) (psi) (L/min) Main Line Pressures and Flows Sprinkler Main Valve 25 W A M Q I 4 1

26 Figure 8 Site A Flow and Pressure (July 30, 2014) 0.4 Bed 10 Flow and Pressure 0.4 Bed 12 Flow and Pressure 0.4 Bed 14 Flow and Pressure Flow (L/s) Flow (L/s) Flow (L/s) Pressure (psi) Pressure (psi) Pressure (psi) A Beds July 30, 2014 Pond Lane Way A-10 A-12 A Pump House North Legend Lateral Line (psi) (L/min) (psi) (L/min) (psi) (L/min) (psi) (L/min) (psi) (L/min) Main Line Pressures and Flows Sprinkler Main Valve 26 W A M Q I 4 1

27 Figure 9 Site A Flow and Pressure August 20, Site A - Flow versus Pressure, August 20/ Flow (L/s) Pressure (psi) R10TG - Black Plate R10TG - Green Plate Linear (R10TG - Black Plate) 27 W A M Q I 4 1

28 Figure 10 Site B Flow and Pressure, August & November 2014 Flow (L/S) Flow versus Pressure, Site B - Aug 14, Pressure (psi) Flow (L/s) Flow versus Pressure Site B - Nov 3, Pressure (psi) Road Way Bed 43 Bed 39 Bed 38 Bed Pump House South North (L/min) (psi) (L/min) (psi) (L/min) (psi) (L/min) (psi) Pressures and Flows Legend Start of Each bed Sprinkler Head Main Pump Line 28 W A M Q I 4 1