Comments on Task 3 Report by EDAW
by David R. Dawdy
September 5, 2003
by David R. Dawdy
September 5, 2003
My comments will be directed toward two major points:
1. The quantity of water contributing to Lake Merced andTo comment first on the water contributing to Lake Merced:
2. The water quality of Lake Merced.
On page 13 it is stated “...ground-water pumping from the deep aquifer...is likely to be the cause of a steep decline in deep aquifer ground-water elevations...” Is there any doubyt that pumping lowers groundwater elevations? This should state that pumping is the cause of the decline.
Further “...it is likely that ground-water in the shallow aquifer currently flows southwestward in response to perennially depressed ground-water levels caused by withdrawals from the deep aquifer.” This is a demonstration of the direct relation of shallow groundwater levels to deep aquifer pumping. The problem is to correctly assess the vertical permeability through the aquitard.
On page 14, 26.4 inches per year of evaporation occurs from the lake, There are about 20 inches per year of rainfall on the lake. Thus, rainfall on the lake is more than offset by evaporation. The net additions to the lake are from surface inflow, which is minor because of the installed surface runoff sewer systems. Part of the surface water inflow deficit may be ameliorated by the Vista Grande diversions to the lake. Actually, Vista Grande is planned to take water away from the lake, and now a part is hoped to be restored. However, the major change in input to the lake is groundwater. Paved streets, rainfall on roofs feeding directly into storm water sewers, and paved over parking lots and front yards in the Sunset district feed water into the storm drains instead of into the groundwater through percolation. This is partly offset by those who maintain green lawns in the summer, over water, and contribute somewhat to the groundwater table. What can be done to enhance the groundwater contribution to Lake Merced?
1. Stop allowing yards to be paved and used as parking lots for excess family cars. This is particularly true in the Sunset District, as mentioned earlier.To comment second on the quality of the water in Lake Merced:
2. Require the new Stonestown development to install pervious paving materials and prohibit them from connecting roof runoff directly to the storm sewer system. Have this requirement applied to all future developments in the basin.
3. Have the Department of Public Works install experimental pervious paving material whenever they repave streets in the Westside Basin. Analyze these experiments and work toward using pervious materials for all streets as they are upgraded.
4. Develop a program for disconnecting roof drains from the storm sewer system and have the runoff flow over yards rather than paved surfaces, which in turn drain into the storm sewers.
5. Develop a grey-water program which uses part of the water now going to the ocean outfall to be used on lawns and gardens, part of which will recharge the groundwater system.
Actually, Lake Merced is quite clean, in general, as shown in Table 2-2 by the physical constituents of the water . Even the nutrients are much lowere than might be expected, as shown by Table 2.3. The shallow groundwater flowing into the lake contains some 30 ppm of nitrate, but little phosphate as shown in Table 4-13. Thus the inflow is phosphorus deficient, and nitrate rich. The lake, as shown in Table 2-3 has the same phosphorus load as the shallow groundwater, but the same phosphorus load. Apparently, the nitrates are taken up by the algae. One way to improve the quality of the water of Lake Merced is to remove the algae.How can that be done? There are at least two means:
1. Have an algae harvesting program with mechanical means in the lake. The removed algae can be composted and made into something useful rather than harmful.As mentioned, the groundwater wells in the shallow aquifer sampled nitrates much higher than the lake. Is there a trend from the wells north of the lake to the wells south of the lake. For example, wells 4, 7, 9, and the Zoo should be little affected by any lake sequestering of algae. The Harding Park well, Well 2, is between North and South Lakes. Is there a decreased nitrate load in the groundwater in Well 2. Wells 1 and 3 should be more affected by any uptake of nitrate by the lake. Are they lower in nitrates than the wells upgradient? If not, why not? The flow of groundwater into the lake should be estimated, and the load of nutrients contributed by the groundwater computed. This would give some order of magnitude estimate whether removal of algae would greatly improve the quality of the lake.
2. Use the Vista Grande facility to remove algae when there is no storm water flowing. This would be most of the year. I assume it will operate by centrifugal force, and the algae can be separated from the clear water, and the clear water returned to the lake. Once again, the algae can be composted and turned into something useful. The Vista Grande equipment is a sunk cost which will be unused some 200 or more days each year. This will make the equipment more cost effective is those 200 days can be used to remove algae from the lake.
The algae are about 2,000 ppm in the lake. (It is not stated at what depth algae measurements are made, but a Secchi disk reading of an average of 1.5 feet means that most of the algae are in the upper foot of the lake, in order to have sunlight.) For about a foot of depth with 62.4 pounds per cubic foot for water and 300 acres for the lake means there are 62.4 x 2,000 x 300 x 43560 / 2,000 x 1,000,000 = about 815 tons of algae to remove to make that top foot clear of algae. This would not be a one-time job. However, removal of part of the algae would improve the quality. Any removal of the algae would leave less algae to die off and add to the organic matter at the bottom of the lake, which causes the anoxic condition at depth.
3. As the lake rises there are sources of nutrients which would be inundated and add to the nutrient load of the lake. Those sources, grasses, bushes, and rushes, should be removed prior to raising the lake level. In the area where groundwater fluctuations occur annually brush should be controlled so as not to add nutrients to the lake.
4. There should be more frequent volunteer cleanup days around the perimeter of the lake to keep potential trash and nutrients out of the lake.
Table 4-12 on page 44 lists depth and screen interval for the monitoring network wells. However, Table 4-13 shows average water quality data. The base data for the averages would shed light on some of the questions asked above.
On page 31, in Table 4-1, the hydraulic conductivity is given for the shallow aquifer and the aquitard, but only transmissivity is given for the lower aquifer. What is the hydraulic conductivity assumed for the lower aquifer, and the thicknesses used to obtain the range in transmissivities? The same parameters should be shown for each aquifer so that some order of magnitude of the assumptions made in the model can be seen.
On page 42, Table 4-8, the E. Coli counts are given for Vista Grande water. They are high. However, the loads may be low. How much water is predicted to be available from Vista Grande in an average year, and how does the load of E. Coli compare with the load already in the lake, which has much lower counts but a much greater volume? That comparison would indicate the magnitude of the problem introduced by Vista Grande E. Coli introduction into the lake.
On page 57 the introduction of pumped groundwater into the lake is discussed. This may be a viable option if the wells are properly situated and if the true relation between the upper aquifer and the lower aquifer is known. The aquitard between the lower aquifer, which would be pumped, and the shallow aquifer, which supplies Lake Merced and of which Lake Merced is a part, is not a uniform, homogeneous member. It was laid down in a marine environment, and therefore is lenticular. In addition, it thins out landward, apparently disappearing in the vicinity of San Bruno Mountain. This can be approximately modeled. As stated, the interaction must be modeled so that it is known how water is recycled from groundwater to lake and back to groundwater over a period of time if this approach is to be implemented.
