Removing Color from an Unstripped Reservoir
How Problem Is Handled by Hartford Water Department—Physical Characteristics of Reservoir—Charts and Tables Explaining Matter
THE writings of Mr. Saville are always filled with practical suggestions and this paper is no exception to the rule. It will be found to contain much that will be of help to other water works men who are compelled to meet the problems of which the author treats:
Although the matter is often one of importance in water supply work, very little information is available on the subject of color of water phenomena in new reservoirs from which the soil has not been stripped prior to filling. The following facts are presented regarding conditions in the Nepaug and other reservoirs of the Hartford, Conn., water supply system, as a contribution, without argument and merely for the purpose of putting them on record, with the hope that they may be helpful to another charged with the solution of the reservoir problem.
There are seven reservoirs of the Hartford water supply system from which the following data have been collected, and for convenience in discussion they may be separated into two groups.
Group 1 comprises reservoirs 1 to 6 inclusive lying on the easterly slope of Talcott Mountain, so-called, which is a low ridge of basalt, protruding through the triassic shale of the Connecticut valley, on the west side of the river of the same name, and located about eight miles west of Hartford. With the exception of reservoir 4 the watersheds contain almost no swampy area. The slopes are quick and covered mostly with second growth hardwood with a considerable sprinkling of hemlock and pine. The entering streams are small, and during the summer months the beds are mostly dry. The character of the soil cover of the rock is disintegrated shale and glacial till, covered with humus, and small springs only are to be found, which often cease to flow during the dryer months. From this it is apparent that the water in the reservoirs is almost entirely that stored during the period of heavy runoff.
Characteristics of Reservoirs
The reservoirs were constructed by dams across the stream of the main valley. Reservoirs 2 and 6 being formed by long side-bill bankments. The bottoms were cleared of timber and brush, but practically little or no attention given to removing or covering soil or small deposits of muck except in limited areas. The data here presented for group 1 have been collected mostly from the four reservoirs listed below;
The characteristics of reservoir No. 4 are entirely different from those of the others; here the brook flowing through a rather broad and level valley was dammed, and later in order to increase the tributary area a low divide was cut through and about 2.1 square miles added. The greater part of this additional area was swampy, and while the additional quantity of water was not in proportion to that from other areas, the quality was much inferior to that of other reservoirs, its color being so high at times as to cause it to be designated “Brandy Brook.”
The average monthly colors, rainfall and reservoir depletion for the nine years. 1915-1923 inclusive, are shown on Diagrams I (Acc. D-1620) and II (Acc. 19-1621). In Table 1 is a comparison of the mean, maximum and minimum colors by months for the nine years of record (1915-1923) for reservoirs 2, 4 and 9.
The physical characteristics of reservoirs 2 and 6 are so similar that some other conditions must be found to explain the differences in color of the stored water, as both of these reservoirs are now well within the limits of what may be be considered old reservoirs with their bottoms in stable condition. Such other conditions may be found in the storage capacity and in the area of water surface exposed to wind and -un action. While the drainage area of reservoir 9 is about double that of No. 2. the area of water surface is four times and to capacity is respectively 3.4 and 2.75 times that of No. 2. It is also probable that the difference in depth of the two reservoirs is a factor to be reckoned with, that of No. 2 being about 46 per cent, more than No. 6. with consequent lessening effect of wind action as regards total circulation of the water in the reservoir.
Reservoir Group 2
Group 2 includes the Nepaug reservoir and its 32 square miles of tributary watershed, which is located about 16 miles northwest of Hartford on a branch of the Farmington River, which it enters about a mile north of the village of Collinsville. The basin is roughly oval, with major and minor axes about 8.5 miles and 4.0 miles respectively, the former lying about northwest and southeast.
The lower part of the basin is underlain with Hartland schist, while the Becket gneiss is the country rock of the upper portion. The whole area was well worked over by the glaciers, as evidenced by the abundance of eskers. kanes and lacustrine deposits in the valleys and the exposed ledge outcrops in the higher portions. The surface topography is somewhat hilly, varying from an elevation of approximately 500 feet to 1.000 feet above sea level. The main Nepaug stream has an average fall of about 1 in 10. The surface of the basin is about equally divided between second growth forest and cultivated or pasture land. The swamp area is very small compared with the whole watershead area, and there are few deposits that might produce unfavorable color conditions. Due to extensive grave! and sand deposits, some of the small streams have a clear, low colored water running nearly all of the time and during periods of low runoff the streams are excellent in appearance, a small area only being most productive of high color.
Formation of Nepaug Reservoir
The Nepaug reservoir was formed by building dams across two streams, the Nepaug River and Phelps brook. By this means the surface of the ponded water was raised sufficiently high to top the low divide between the two streams and thereafter a single pond resulted.
Previous to its filling the basin contained 47 farms, the level bottom lands along the main stream being used for cultivation. while the steeper side slopes were mostly used for pasturage, with here and there patches of timber and second growth brush. The area was approximately divided into 343 acres of cultivated land, 168 acres timberland, 205 acres sprouts, 122 acres pasture, and 12 acres swamp.
In the construction of the reservoir all buildings were removed; all trees and brush were cut to ground level and the shores grubbed of all roots tor ten feet below and five feet above high water. Before final filling the whole reservoir bottom was gone over with scythe and rake, all vegetable growth cut and burned, and the basin left in as clean and clear a condition possible. One small area of peat deposit perhaps 2 or 3 acres in extent was noted but was not treated but for the most part the reservoir bottom was underlaid with sand and gravel covered with a sandy subsoil underlying a rather thin top soil, bare where formerly cultivated, and grass grown elsewhere.
(Continued on page 67)
Removing Color from Unstripped Reservoir
(Continued from page 60)
The total capacity of the reservoir is 9,500,000,000 gallons, its average depth 34 feet with a maximum depth of 97 feet, near the Nepaug dam. There are few areas of shallow flowage. the largest single span being at the north and where there are about 10 acres with a maximum depth of 6 feet. Over this area the soil was stripped off to the gravel.
Four principal streams enter the reservoir. Clear brook, a small tributary, coming almost entirely over open sand bank country; Phelps brook, with a somewhat larger tributary area, draining a mixed country, the lower portion steep and rocky with forest cover, the upper part with much open farm land and some swamps; the Nepaug River, the main stream, draining about three-quarters of the whole basin, which is ordinary New England country area—pasturage, open farm land and forest growth and some bare rocky and sandy areas, but little swamp land that would be of detriment to the runoff; the fourth is the Collinsville reservoir stream, drawing about 0.4 square mile, swampy at its lower end but heavily timbered on steep rock or gravel slopes in its upper portion.
Table 2—Nepaug River, 1917-24, Rainfall, Run-off and Colors
Table 3—Phelps Brook, 1917-24, Rainfall, Run-off and Colors
Color Characteristics of Nepaug
The color characteristics of the Nepaug reservoir and its streams have been carefully observed since work first began. The study began in 1912 with the streams running in their natural channels prior to the beginning of construction work and has been continued to the completed and seasoned reservoir in 1924. See tables 2 to 5 inclusive.
In 1912 the average color of the Nepaug River was 31 with a maximum color of 82; in 1913 the colors wore respectively 27 and 58. For Phelps brook the average color observed in 1915 was 48. with a maximum of 83. From 1917 on both color and runoff observations usually have been taken weekly at many regular stations in the reservoir and on the principal tributary streams.
Ascertaining the Average Color
For purposes of ascertaining the average color of the water delivered into the reservoir the following weighting, proportionate to area, was used with the data from the respective supplies:
The color of the effluent water in the following discussion is that observed in samples taken in the Nepaug reservoir near Phelps brook dam.
(Continued on page 76)
Table 4—Clear Brook, 1917-24, Rainfall, Run-off and Colors
Table 5—Collinsville Reservoir, Tabulation of Colors of Raw Water
Removing Color from Unstripped Reservoir
(Continued from page 67)
The color of the influent water is that obtained by the combination explained above, using the colors of water from the principal entering streams together with zero color for rainfall on the reservoir surface. The annual averages are given in table 6, and are shown also on Diagram III.
Diagram III shows the progressive decrease in color of the stored water at surface, middepth and bottom.
Table 6—Average Annual Colors, Nepaug Reservoir and Tributary Streams (Reservoir Filled in 1917)
“Weighted, and using color “O” for water falling on reservoir surface. Average would mean nothing because of changing conditions of reservoir
Diagrams IV shows several interesting conditions derived from the colors of the several entering streams and weighted as stated above. While the effluent water has approximately the same color in both years, the color of the influent water depends very greatly on the amount of stream flow. In 1922 with a comparatively high runoff during the entire season the color of the influent water was considerably above that of the effluent, while in 1923 during the period of low runoff, June to September, the color of the influent was only slightly above that of the effluent. Again, note that during the winter season with ice cover on the reservoir and little action of sun or wind the effluent water color is measurably above that of the incoming water.
On Diagram V is shown changes in color of the stored water since the reservoir was put in commission.
The lines A-15, C-D, 15-F respectively indicate mean maximum and minimum color range. The full and dotted horizontal lines indicate respectively the mean colors of the reservoir water and that of the Nepaug River, which supplies about two-thirds of the entering water.
Diagram VI shows the relative changes of color in the reservoir at different depths at the same time of observation and at different periods after the reservoir was filled (which was in the fall of 1917). Attention is particularly directed to the improvement in color of bottom water between Septemper, 1920, and the same time in 1923.
As a record of conditions of percentage of saturation temperature and color of top, middle and bottom water by months, Diagram VII has been prepared for the year 1920, and for purposes of comparison records for May, June and July, 1924, have been plotted on the same chart.
Color Effect of Small Stream
The color effect of comparatively small streams entering a large reservoir, be they excessively low or high, seem of relatively small importance, as it is probable that zero rainfall on the reservoir itself or the discharge from springs directly into the reservoir will neutralize the high colors from small streams flowing from swampy areas. While it is eminently desirable to improve watersheds by swamp drainage and so reduce color of the reservoir water if it can be done to an appreciable extent, it is also desirable to study closely local influent conditions and decide if the result to be attained is worth the cost of the effort.
The following data are given for periods in which excessively high colors were observed in Phelps brook:
The Phelps brook watershed of 2.9 square miles drains a swampy area of perhaps a quarter of its entire drainage area. Clear brook with a drainage area of about 1.05 square miles is entirely a glacial gravel and sand country, while the Nepaug reservoir surface surface is 1.33 square miles in area with an addition of about 2.25 square miles steep slopes discharging quickly into the reservoir. The effluent colors in the reservoir are taken at the outlet chamber about 1.25 miles almost directly east of Phelps brook outlet and in the path of prevailing winds.
Rainfall Record of Whole Basin
The following rainfall records (inches) practically cover the whole drainage basin:
The April 6-7, 1924, storm of 4.03 inches plus some snow accumulation produced one of the greatest drainage basin disturbances, of recent years, and probably has not been exceeded in the past 25 years. The following colors were observed immediately before the storm commencd, during the height of the discharge and just after the discharge had returned to normal.
There was no appreciable increase in color of the Nepaug reservoir water in any of the weekly determinations from April 2 to May 28. For colors at different depths the following determinations are given, for two localities where deepest water is found:
Color Diminishes with Time
From what has gone before it appears that in a reservoir similar to that of the Nepaug reservoir of the Hartford water supply system and with similar conditions of inflow, after five or six years the reservoir gets conditioned, as it were, although for several years longer there will be a gradual decrease in the color content of the stored water.
The data presented in this study substantiate the conclusions of that master of water supply phenomena, Frederick P. Stearns, that when an unstripped reservoir is first filled the water acquires after standing for a time a considerable amount of color, chiefly from vegetable matters which have been flooded. Some of the color taken up is removed by the bleaching process which goes on in reservoirs where the water is stored for a long time. In the early years more color is taken up than is removed and the residual color is greater than the average color of the inflowing streams. As the years pass, the amount of coloring matter acquired from the bottom and sides of the reservoir diminishes. Meanwhile the bleaching process is continually going on, so that after a term of years the result of storage is a decided diminution of the color of the water which enters the reservoir.
In regard to this color diminution in unstripped reservoirs, the period of attaining what may be called a normal color is somewhat uncertain due to local conditions, but the following are the results of study of some data which were made at the time when the probable future conditions in the Nepaug reservoir were being considered:
(From paper read before the annual convention of the New England Water Works Association.)