Results of Water Waste Survey in City of Detroit
Important Work in Conservation of Water — Underground and Other Leaks Located and Stopped—Defective Plumbing an Important Element of Waste — Efficiency of Meterage
GEORGE H. FENKELL, the author of this paper, was born at Chagrin Falls, Ohio, February, 4, 1873. He became civil engineer of the commissioners of water works, Erie, Pa., in 1902, and civil engineer to the board of water commissioners, Detroit, Mich., in 1908. In 1913 he was appointed commissioner of public works of the city of Detroit by the mayor, and in 1918 resigned that position to become general manager of the board of water commissioners, Detroit.
The information that has been gathered on the consumption and waste of water in Detroit is of interest for several reasons.
- A water works system was installed in Detroit nearly 100 years ago, and pipes laid 80 years ago are still in service.
- Ninety-eight per cent, of all active service connections are now metered.
- A pitometer survey covering the entire city is nearing completion.
- As Detroit is served with a direct pressure system, hourly records of consumption are available.
- Detroit is the largest city in point of population in the United States—and the writer believes in the world—having nearly all of its connections metered.
- As the quantity of water available from wells has been negligible, the people of Detroit have been at all times compelled to obtain their entire supply for domestic purposes from the distribution system of the water works.
Description and History
The city of Detroit is located on the north and west side of the Detroit River on land very uniform in contour, the highest point within its limits being 65 ft. above the surface of the river. The water system was installed by a private company in 1827, and in 1836 this was purchased by the city. When Detroit was but a hamlet, the settlers were unable to secure water from wells and were compelled to obtain their supply from the Detroit River, because of the nature of the material underlying Detroit and vicinity. This deposit has a thickness in Detroit of over 100 ft., and is composed of a compact bluish and unstratified clay, charged with pebbles and boulders of glacial origin, and generally impervious to the flow of water.
Beginning with the first government census, the population of the city has increased steadily as follows: 1810, 707; 1820. 1,442; 1830, 2,222; 1840, 9,192; 1850, 21,019. From 1850 to the present, the population is shown graphically in chart No. 1, on page 206, and estimates recently made place the population at 1,020,700.
While the use of most of the pipe laid prior to 1853 has been discontinued, there still remains some that has been in service eighty years or more, as for example, a section of the ten-inch cast iron pipe in Jefferson Ave., laid in 1839. In 1861. or 60 years ago, the distribution system contained 39 miles of cast iron pipe, and it is estimated that about onehalf of this is now in use. At present the distribution system contains 1480 miles of cast iron pipe in sizes from 4 inch to 48 inch in diameter, and 3 miles of 48 inch steel pipe.
In 1886 the use of a reservoir was discontinued, and since that date the system has been operated as a direct pressure system. The first Venturi meter was set in a discharge main in 1902. and since 1913 all water pumper has been metered at the pumping station, and before 1913 allowance was made for pump slippage. The service connections are of lead and cast iron.
Systematic Metering Begun in 1889
Systematic metering was begun in 1889. when 194 meters were installed, and by the end of 1890 this number was increased to 856. The accompanying chart shows that approximately 10 per cent, of all services were metered by 1897, and this ratio of the number of meters to the number of service connections continued fairly steady until 1912.
It will be seen by referring to chart No. 1 that in 1882 the per capita consumption was 120 gals, per day and in 1887 was 208. This phenomenal increase occasioned much discussion for several years. The marked decrease from 1887 to 1897 was due to at least three causes, to wit: the use of meters, which were at the time credited with all of the saving, although the number set was small: careful inspection of premises; and the installation of by-passes on the vertical compound double acting pumps. The introduction of these by-passes permitted the capacity of the two 24 mil. gal. pumps to be changed at will to 12 mil. gal. pumps. The by-passes made it possible to eliminate the waste of large quantities of water which were allowed to flow into the suction wells at times when the capacity of these pumps working at minimum speed was greater than needed to supply the city.
In Detroit, general metering was begun in 1913. and on July 1, 1920, 97.4 per cent, of all active service connections were covered. The greatest number of meters set in any month was 5,345 in October, 1918.
The average water pressure as determined from 39 pressure gauges generally located in fire engine houses, in various parts of the city and read hourly is 31.5 lbs., the maximum pressure is 45 lbs., and the minimum pressure is 22 lbs.
Consumption and Waste
The items that enter here may be divided and classified in various ways. A table has been prepared in the way that was most convenient for use in collecting the data, and the following explanations and remarks relate to this table.
In computing the quantities based on meter measurement, an allowance of 5 per cent, has been made for under-registration of meters.
The first item covers 131,692 metered domestic accounts. This number was arrived at by subtracting the number of commercial accounts, as estimated from a careful examination of the city directory, from the total number of metered accounts where meters were one and one-half inch or smaller.
The water furnished 2,985 domestic flat rate accounts is estimated by allowing one thousand gallons for every 5 cents of revenue. The meter rate in force is on a quantity basis and is one dollar for the first thousand cubic feet, fifty cents per thousand for the next three thousand cubic feet, and thirty-five cents per thousand for all additional.
On the 30th of last June, there were 164,779 service connections of all sizes in the city, and of these 145,053 were in use. On May 1, 1921, it was estimated that the area covered by the water waste survey contained about four-fifths of the total population of the city. The total number of service leaks found by this survey in this area was 447, wasting 4,359,000 gals, per day, or approximately 10,000 gals, for each service per day. All of the water wasted by the service leaks just mentioned was unmetered. By pitometer measurement 116 of these were found to be wasting 2,704,000 gals, per day, or about 23,300 gals. each. The remaining 331 were small service leaks which were located in connection with the inspection of the houses for leaks and are estimated to have averaged 5,000 gals, per day each.
The population of Detroit in 1920, as determined by the United States census, was 993,739, and at present it somewhat exceeds one million. For the purpose of this paper, a population of one million has been used, and it is believed that no serious error has been introduced by the use of this figure.
All services having meters 2 in. or larger are considered to supply factories and other commercial users, 1,219 in all. This number includes some large apartment houses, but it is believed that the error introduced is small.
The number of stores and other commercial accounts which is estimated to be 7,416, was arrived at by an examination of the city directory as previously explained. The water used for building purposes is found by allowing 1,000 gals, for every five cents revenue in the same manner as with the domestic flat rate accounts.
All water supplied to villages and factories outside the city limits is metered and the number of domestic consumers so supplied is negligible.
Water Supplied for Municipal Purposes
The estimate of the total amount of water furnished for municipal purposes is 17.44 gallons per capita per day, or 12 per cent, of the total pumpage, or 13 per cent, of the total quantity of water consumed within the city limits. Included in this is 0.39 gals, furnished free to charitable institutions and 0.95 gals, at one-half rate to other charitable institutions. The department of water supply receives no revenue for 12 per cent, of the water supplied the city. Flooding trenches is credited with 7 cu. ft. per lineal foot for 200 miles of trench. The estimates for paving, sprinkling, and flushing streets, and for flushing sewers are based on data furnished by the department of public works. Four of the 42 engine houses are metered. The water used in 15 ladder houses is estimated. Five of the eleven police buildings are metered, and the supply for five parks, including Belle Isle, is metered. There are 12 unmetered park fountains, and the Pitometer Company found the average daily consumption for each of these to be 45,000 gals. There are 150 continuous flow sanitary drinking fountains on the streets of the city, each of which is estimated to use 1,100 gals, per day, and 150 house fountains, each using, as determined by measurement. 18,000 gals, a day.
(Continued on page 216)
(Continued from page 206)
The public welfare commission during the winter months makes use of eight skating ponds, and it is
estimated that each has a capacity of 20,000 cu. ft., and that it is filled four times a season. All of the 14 public libraries are metered, 123 of the 161 public schools are metered, and the water used in 75 private schools is estimated. Of 35 other public buildings, 8 are metered. There are 32 charitable institutions receiving free water, and 30 on a half rate basis, all of which are metered.
A careful estimate shows that 30,644,000 gals, of water were used in extinguishing fires during the year ending June 30th, 1917, and it has been assumed in preparing this estimate that there has been an increase of 40 per cent, since that time in the use of water for this purpose.
Leaks Uncovered and Stopped
Up to May 1, 1921, the Pitometer Company has located, uncovered and stopped leaks in mains amounting to 2,533,000 gals, per day. During the past year there were 75 breaks in mains, and it is estimated that this loss amounted to 32 mil. gals. Of these broken mains, 5 were 24 in. in diameter or larger. Great care has been used by the department in laying large mains, and the following test is mentioned to show the result of careful workmanship and good inspection. In 1909 a section of 48in. cast iron pipe 14,907 ft. long, containing 1,360 48-in. joints; seven 42-in. joints; two 36-in. joint, and thirty-eight 12-in. joints was tested after closing two 48-in. valves; one 42-in. valve; one 36-in. valve, and fifteen 12-in. valves, by supplying the entire line through a 1-in. meter for a period of 20 hours, and the quantity so supplied amounted to 690 cu. ft. which was at the rate of 258 gals per hour.
The quantity used for flushing mains includes that used for filling new mains and making blow-offs of dead ends, and is estimated to be 500 gals, per minute through the year.
The consumption figures given herein are derived from meter measurements.
The Consumption of Water in Detroit
The figures in the preceding table have been presented without attempting to have the amount accounted for equal the total consumption. It is known that considerable amounts are used that are not ineluded in the table. For example: extinguishing fires in coal piles; use in water works park not included in other parks; street sprinkling wagons operated in business districts; and sprinkling street car tracks by street railway.
(Continued on page 222)
(Continued from page 216)
Some of the difference can also be accounted for because of the industrial depression during the past few months, for the consumption in September last was 159 gals., while in March it was reduced to 131. This affects the table, because the metered consumption for any particular month as determined by meter readings taken throughout the city includes water actually used two to five months previously, while the Venturi meters at the pumping station determine the consumption from day to day.
According to estimates presented, the daily per capita consumption for domestic use is 56.05 gals., and an attempt will be made to analyze these figures, by separating this quantity into several items.
It is difficult to determine the domestic requirements by actual measurement, although attempts have been made by the Pitometer Company.
In the survey covering the period from September, 1919, to May, 1921, an examination was made of all properties where evidence seemed to show a waste. In only 50 cases was water found running to prevent freezing, amounting to 101,632 gals, per day. As this survey continued through all seasons of the year and only covered a small area during the season of frost, only a small part of the waste to prevent freezing was detected. The amount of such waste naturally depends upon the temperature, beginning at 32 degrees F. early in the winter and 40 degrees F. late in the winter, increasing rapidly as the ternperature drops below these points
Industrial depression during the winter of 1920-21 reduced the consumption of water, particularly for manufacturing and railroad purposes, to such an extent, and the mildness of the winter reduced the need for putting into effect anti-freezing methods to such a degree, that the consumption tables give little information on which to base estimates of the quantity of water used to prevent the freezing of plumbing. It has been found by measurement that the average waste from one tap to prevent freezing is 2,000 gals, per day. When one tap in 100 is allowed to flow for three months, the average daily waste for the year will amount to 0.75 gals, per capita; likewise when one in 50 is allowed to waste, the average is 1.5 gals.; one in 25, the waste is three gals., and one in ten the waste is 7.5 gals. Before Detroit was metered, this waste could be expected to equal or exceed the last figure, but with meters it will probably exceed 0.75 gals., and be less than three gals. Because of the mildness of the winter of 1920-21, the minimum amount, 0.75 gallons, will be used in this estimate.
The summer of 1920 was cool and comparatively little water was used for sprinkling. From an examination of the charts made in connection with the water waste survey, there is but little evidence of an increased flow during the evening hours, and it is not possible from these to make an accurate determination. If it is assumed that two-thirds of the services are used at some time for sprinkling purposes, and that each service is used for one-half hour every second day through a period of three months, the yearly consumption will amount to 1.8 gals, per capita. It seems probable that with all services metered, water for sprinkling will be used at the rate of between 0.9 gals., and 2.5 gals, per capita, and for the season under consideration the former quantity will be used.
(Continued on gage 230)
(Continued from page 222)
The method of arriving at the estimate of the leaks in service connections, 5.45 gallons, was determined by the water waste survey and was explained on a previous page.
Leaks Through Defective Plumbing
The estimate of 22.33 gals, per capita per day for leakage through defective plumbing fixtures has been made in the following manner:
Estimate of the Daily Per Capita Consumption of Water Used at Night Rate for Public Purposes
In making a study of the use of water for domestic purposes, a considerable amount of data was collected in April of this year in an endeavor to determine the minimum night flow in districts of varying character.
While the results are of a conflicting nature, they are mentioned here as they have a bearing on this subject.
One block in the poor class foreign district was measured, first with a pitometer, and later with a 5/8-in. disc meter. It was found that the rate from midnight to 5 A. M. on Adelaide St. between St. Antoine and Hastings, population 242, 8 leaking faucets, 9 leaking toilets and 2 other defective fixtures, was 724 gallons per hour and after making an allowance for leaks in fixtures, this was reduced to 339 gals, per hour, which is the equivalent of a net night rate per capita of 33.6 gals, for this block.
The consumption of water in two high class residence streets was measured as follows: Boston
Boulevard between Woodward and Third Streets, population 83, 12 service connections. 3 premises equipped with Brooks lawn sprinkling system, with a total of 1,120 spray nozzles. This block when measured with a 3/4-in. disc meter from midnight to 5 A. M. gave a night flow at the rate of 64.4 gals, per capita per 24 hours. No opportunity was given for the inspection of premises. The north side of Chicago Boulevard between Woodward Avenue and Third Street, with 14 services supplying 83 people, under the same condition gave a per capita of 37.2 gallons.
(NOTE—Excerpts from paper read before the Cleveland convention of the American Water Works Association. To be continued.)