Problem of Pollution in Large Water Tunnels

Problem of Pollution in Large Water Tunnels

Chicago’s Experiences in Its Seven Large Gravity Tunnels—Some Lake and Other Land Tunnels—Matter of Seepage and Leakage

CHICAGO’S experience with its large water tunnels and the danger in some instances of pollution through seepage or leakage forms an interesting subject for consideration and Mr. Gorman has handled the matter both painstakingly and exhaustively:

For supplying water from the six inlet cribs in Lake Michigan to its ten municipal pumping stations, where it is chlorinated before being delivered to the distributing system, the city of Chicago has in use seven gravity tunnel systems comprising 58.8 miles of tunnel. About 22 miles are under the lake, the remaining 35 odd miles being land tunnels. Of the latter about 25 miles are under city streets and approximately 11 miles under private property. The tunnels are at various depths, ranging from about 40 feet below city datum (mean lake level) to approximately 150 feet. The earlier tunnels were dug through clay or clayey sand and were built of brick. The latter ones are of concrete construction dug through rock.

Map of the Chicago Water Tunnel System, Showing Lake and Land Shafts and Tunnels

Two tunnel systems, the Crosstown and the Old Polk street tunnels have been abandoned. These crossed the heart of the city under private property and were injured by piling and deep foundation construction. They have been filled up to a considerable extent and have been plugged off at convenient places by concrete cores. The New Polk street tunnel system is crossed by the former tunnel at two points, which are being investigated in connection with an abnormal degree of contamination of the untreated water supplied to the Harrison street station.

Data Concerning Water Tunnel Systems in Chicago

Tunnel Shafts

In constructing these tunnels, working shafts usually about 10 feet in diameter were sunk.

Lake Shafts: Those constructed under the lake and protected by temporary cribs during the digging of the tunnels were sealed off at or about the tunnel crown and back-filled to the bed of the lake. The crib was then demolished. There were 12 of these lake shafts sunk in connection with the Chiago tunnel systems. On the 68th street tunnel system alone there were four lake shafts, this number being necessary on account of construction difficulties in tunneling through sandy sub-strata.

Land Shafts: In the earlier tunnel work land shafts were made of brick; in the later tunnels of concrete. Practice has differed with regard to the maintenance of these shafts. In the 70s and 80s they were considered as a reserve source for water for fire purposes. In fact, on the Cross-town tunnel, which has since been abandoned, smaller shafts were especially constructed at intermediate points to give a reserve supply of water for fire protection. In a few instances private companies using large quantities of water were permitted to make connections to these shafts. The danger of this practice will be discussed elsewhere. The policy in recent years has been to crown over working shafts at the tunnel by a reinforced concrete slab several feet thick and to back-fill with the material removed from the tunnels. Shafts in which gates are set, and others deemed important for tunnel maintenance and operation, have been left open. In several of the older shafts not considered important a concrete slab crown was constructed at a point above the water line and the shaft filled with sand, gravel or cinders.

As a rule shafts were sunk in public property, usually in a street. Undoubtedly due consideration was given to the relative location of sewers with reference to shafts, but with the rapid growth of the city many changes in underground systems in streets have been required, some of which constitute a potential danger to the water system. However, it does not appear that the existence of a sewer in a street forbade the sinking of a tunnel shaft in the immediate vicinity, or that the presence of a tunnel shaft prohibited the laying of a sewer in the same street. The city engineer has, in several instances, insisted that the alignment of sewers be altered to assure reasonable clearance around tunnel shafts. As a considerable portion of the sewer system of Chicago has become overtaxed one can realize the potential danger which a sewer flowing under pressure constitutes to a nearby tunnel shaft in which the gravity pressure is less than the head, due to lake level.

The method of protecting the shafts at the surface has depended upon the importance of the shaft. In Rate shafts and other pivotal ones heavy cast iron covers have been used, the joints between the cover and the frame usually being sealed with cement grout. In the case of shafts of lesser importance the cover was crowned with concrete and the ground or roadway laid over same.

Showing Danger of Contamination Through Private Cross-Connection

Water Sampling From Tunnel Shafts

Early in 1924, in accordance with a co-operative arrangement between the departments of health and public works in C hicago, a new program for supervision and control over the public water supply was inaugurated. To permit of a more satisfactory supervision over the chlorination at the pumping stations twelve sampling pumps were installed at representative points in the tunnel system, which provided for the collection of samples of water from the tunnels for bacterial analyses, The selection of the fixation of these sampling points was made with a view that samples analyzed might show:

  1. Any contamination of the water in the lake tunnels between the Cribs and the lake pumping stations.
  2. Any contamination ot the water in the land tunnels between the lake front pumping stations and other stations located inland.

These pumps are of the ordinary farm type and are set up over a shaft with the suction line extending down to the center of the tunnel. W hen located out of doors they are protected by a steel housing.

It was not until late summer that all these sampling pumps were installed, but the information supplied by the analyses of samples collected has been most useful and has given a remarkable picture of conditions which could not have been otherwise detected. The analyses of these samples directed our attention to and enabled us to trace and take measures to remedy dangerous sources of contamination found to exist. The results have clearly demonstrated the value and importance of making routine bacterial analyses of water samples from all representative points in the public water system, and clearly emphasized the importance of extended laboratory control over the quality of a water from its source to the consumer’s tap. With the information obtained from these bacteriological surveys, City Engineer John Ericson, has arranged for a complete physical survey of all suspected tunnel shafts. In the 1925 approprigtion the City council authorized an expenditure of $50,000 to properly protect the tunnels in this city.

Investigation of Pollution of Lake Tunnels

The circumstances surrounding the outbreak of typhoid fever in November and December, 1923, in the water district supplied from the 68th street crib strongly suggested the possibility of intermediate contamination of the water passing through the tunnels from the crib to the pumping station. As the history of the construction of the 68th street tunnel system showed two serious cave-ins, and the necessity of abandoning a section of the tunnel, an investigation of this tunnel system was arranged for early in 1924. At this time a series of water samples for bacterial analyses was collected from shore shafts on each branch of the tunnel to determine, if possible, any marked difference in the quality of the water. At the same time samples of “raw’’ water from the land tunnel coming from the Dunne crib, which structure is located in the lake within 150 feet of the 68th street crib, were collected. The results of analyses of these samples were not striking, but indicated a better quality of water from the Dunne than from the 68th street tunnel system. As this did not eliminate the factor of possible local contamination at the 68th street crib (since crib tenders live on this structure and not on the Dunne crib) a later series of samples from the cribs as W’ell as the land shafts was collected over a twelve-hour period. Upon analyses these samples failed to indicate any marked increase of pollution between the crib and the shore.

Diagram Showing Crib, Tunnel and Pumping Station

As the 68th street tunnel is now known to leak the lesson learned from these negative results has been that the analyses of even a considerable number of samples over a short period of time may give misleading information. Contrasted with these results we have the daily analyses of samples from the 68th street and Dunne tunnel svstems which defin’tely indicate that the water from the former source was quite consistently more contaminated than from the Dunne system, although the cr*b intakes are served from the same area in Lake Michigan. There were, of course, exceptions during the summer months, hut it is believed that this was due to the fact that during this period the lake water surrounding both intakes was considerably more contaminated than the water between the cribs and the shore which filtered through the bed of the lake in entering tlv 68th street tunnel.

In February, 1925, this tunnel was put out of service and the bureau of engineering made a test for leakage, which gave very positive results. Samples for bacterial analyses collected during this test showed no marked contamination, although the bacterial counts increased considerably when the normal pressure in the tunnel was lowered by pumping. The city engineer has wisely directed that arrangements be made to supply the pumps at the 68lh street station, formerly served by this tunnel, with water from the Dunne crib, after which the bureau of engineering will dewater the tunnel and investigate its condition.

Diagram Illustrating Contamination of Tunnel by Bilgewater Through Shaft

During the year studies were also made of intermediate contamination of the other lake tunnels. No marked evidence of leakage has been obtained.

Investigation of Pollution of Land Tunnels

Pollution of the Green Street Shaft: The results of analyses of water samples from the first three pumps installed in tunnels showed no abnormal conditions beyond somewhat higher bacterial counts the first few days after each pump was put in service. When the pump at Carroll and Elizabeth streets (Central Park R-l) on the south branch of the Northwest land tunnel was first installed the samples showed excessive contamination and attracted attention. A special series of the samples collected while the pump was operated continuously made no improvement in the quality of the water, it was found. The pump was then suspected as being the source of pollution and was removed. The end of the suction pipe had a collection of green slime, but beyond this nothing unusual vvas noted. The pump and suction line were thoroughly rinsed with a solution of calictim hypo-chlorite and re-set. The samples collected continued to be grossly contaminated. An investigation of the tunnel shafts was then begun.

It was noted that the degree of contamination increased after heavy rainfalls. It has been our experience that defective conditions permitting surface contamination of tunnels give magnified bacterial results directly after excessive rainfall rates and this has been taken advantage of in making tunnel investigations. In this connection it has been found that the keeping of extensive meteorological records materially assisted us in water supply control in Chicago.

Diagram Showing How Chlorine is Applied in Suction Well at Pumping Station

(Continued on page 181)

(Continued from page 158)

The tunnel shaft investigation work consisted of a physical inspection of the interior of the shaft for leakage through the wall and the collection of several samples of water at various depths in the shaft until the tunnel was reached. Samples were collected and sterile bacterial bottles set in deep water sampling apparatus and suspended by a cord. It is believed that the samples collected were reasonably representative of the water at the depth sampled.

Although evidences of surface seepage into other shafts in the Northwest land tunnel system were noted, results of analyses of samples from the Green street shaft were the only ones showing gross contamination. This shaft is at the intersection of the branches supplying the Central Park and N. Springfield avenue pumping stations. It has been crowned over with a slab above the w’ater line and has been filled with about four feet of gravel. A hole had been drilled through the shaft for some previous investigation of tunnel velocities and left open so that any seepage which passed through the gravel could enter the shaft through this opening. Evidences of such seepage on the wxst wall were noted. A sample of the slime collected on this wall when washed with sterile w’ater and analyzed was found to be grossly contaminated by B. Coli. The records of the sewer division were consulted and showed that a 24-inch circular brick sewer about 50 years old passed within one foot of the west wall of the shaft, the invert being at a point just about the line of seepage noted.

Arrangements were then made to have the sewer opened, W’hich was done in September. It was found that the main sewer was in good physical condition, being surrounded Inhard clay. There were evidences of seepage around the sew’er, especially toward the shaft w’here several decayed sheeting posts were found. Within 10 feet of the shaft there was a four-inch house drain w’hich w’as cracked and leaking at a point near the main sewer. Four other house drains at distances up to 50 feet from the sewer were located. Two of them w’ere found in the same condition as the nearer one and leaking. It is doubtful if any great quantity of sew-age leaked into the shaft at one time, but as the main sew’er was filled with deposits to one-half its depth and is known to have flowed under pressure following heavy rains it is reasonably certain that sew’age contaminated seepage water surrounded the shaft at such times and probably polluted the w’ater flowing in the tunnel.

After conditions around this shaft w’ere determined the city engineering arranged to have this sewer replaced in the vicinity of the shaft bv cast iron water pipe with leaded joints. The “Y” branches to the house drains were also made of cast iron and the joints leaded.

As this work necessitated the removal of parts of the sew’er and required several days before the entire cast iron pipe section w’as laid it was necessary to take extraordinary precautions to protect the shaft should a heavy rainfall occur and cause the adiacent sew’ers to flood the open ditch. Accordingly arrangements w’ere made to chlorinate the shaft day and night while the repair work was in progress. Chlorine gas w’as delivered directly into the shaft water by means of a three-quarter inch hose attached wdth a special fitting to a gas cylinder. Shortlv after this chlorination began samples from the pump at Carroll and Elizabeth streets showed a marked improvement in bacterial quality. Since this day the w’ater from this source lr>s been reasonably representative of the quality of the water from the Carter H. Harrison crib.

Contamination of New Polk Street Tunnel System

There are two water sampling pumps in this tunnel. One is at Park Row in a land shaft at the junction of a branch to the 14th street pumping station, and the other in the suction well at the Harrison street pumping station, about one and one-half mile westward via the tunnel route. Samples from the former shaft upon analyses showed no unusual condition. The Harrison street sampling pump was installed on September 8, 1924, and immediately the results of analyses of water from this source indicated something radically wrong.

This tunnel was looked upon with suspicion as it has quite a history. It was built in 1907 of concrete and was dug through “clav. In a distance of something over one mile from the Peck Court riiaft at Grant Park to the Harrison street pumping station the main tunnel is crossed over twelve times bv a freight tunnel under the business district of the citv. once bv the abandoned Cross-town tunnel, and is connected into in two places by the Old Polk street tunnel. In addition to this a connecting tunnel between this system and the Blue Island tunnel system is crossed by the Cross-town tunnel.

By way of illustrating the danger of permitting private connections to be made to water tunnel systems, two conditions which were found in the New Polk street tunnel system will he cited:

1. In the basement of a warehouse there is an opening into a valve pit f, a suction line connected to a tunnel shaft. A public service company had in earlier days obtained permission to supply its pumping station with water from this source. While investigating this opening a tire hose was found hanging into the valve pit. At intervals of about one half minute a flow of water through the hose discharged into the valve pit. A further investigation showed that the hose was from a bilge [tump set up to drain a large cavity under the building which had been washed out. due to the underground leakage of water front some unknown source. The contractor who had been employed to repair the leak had installed the bilge pump and stated that he believed he was discharging the bilge water into a sewer catch basin instead of a pit which indirectly connected to a city water tunnel. Although the pit thought to he sealed off from the tunnel by a concrete wall the bilge water did not run into the pit and apparently leaked into the tunnel.

Further investigation showed that a second potentially dangerous condition existed. A brewery had at one time been given permission to draw water from a fire shaft in the Cross-town tunnel through a sixinch private suction line. This shaft was connected to one of the main shafts in the New Polk street tunnel, the gate between same being open, lust above the fire shaft there was a shut-off valve m the former ‘six-inch suction line. Fortunately this had been closed, for had it been otherwise the city might have experienced a serious typhoid fever outbreak in the area supplied by the Harrison street pumping station. An investigation of the basement of the former brewery showed that although the old pumping engines had been removed the six-inch suction line was still open and protruded through the wall of the building at the level of the basement floor. Under the floor of the pump room were two catch basins connected to the sewer, both being open at the top. During periods of heavy rainfall had sewage backed up into these catch basins sufficiently to flood the earth floor of the pump room the open six-inch suction line described aliovc would have served as a drain for the overflow sewage discharging into the fire shaft, thence to the water tunnel if the valve mentioned above had not been closed. 1 be fire shaft has since been filled up by the Bureau of Engineering.

During the course of the investigation of the pollution of the “raw” water at the Harrison street station suspicion has been directed against the Cross-town tunnel which crosses the New Polk street tunnel and its connecting tunnels at two points. 1 o permit of the collection of samples on either side of one of these points the back-fill in two shafts was removed and special holes drilled through the slabs covering the shafts. Results of analyses of samples collected from these shafts showed that the concrete cores in the abandoned tunnel were apparently effective in sealing off this cross connection. However, the second crossing of the abandoned Cross-town tunnel with the New Polk street tunnel is still under suspicion, since it is known that during its construction nearly 50 years ago difficulties were experienced in tunneling in the area concerned. The evidence to date points strongly to this as being the probable source of contamination.

In connection with the investigation of the New Polk street tunnel system, it is planned to study the possibilities of seepage of ground water into the tunnel as it passes under the Chicago River, where penetration of ground water is known to be unusual, due to a poor character of sub-strata. It may be possible that at times the static head of ground water surrounding the tunnel is greater than the outward pressure of the water in the tunnel, because of the lowered head, due to friction losses from six miles flow in a gravity tunnel. If such is the case and the walls of the tunnel are not tight, ground water would be able to seep into the tunnel.

The proximity of the Illinois freight tunnel which crosses over the Polk street tunnel twelve times in one mile, and the possible effect of same will also be studied. On the west side of the Chicago River there is only 4.3 feet difference between the crown of the water tunnel and the invert of the freight tunnel. On the east side this difference is 6.3 feet. However, as the freight tunnel has pumps for the collection of drainage water there is need for very careful investigation of the potentialities for pollution entering the water tunnel. Of course, under normal conditions the freight tunnel would be in most jeopardy since both tunnels are upward of 50 feet below the lake level. However, since there arc private shaft connection? in many large buildings in Chicago to this freight tunnel, in times of extreme emergency, such as fire, this tunnel may become flooded by water or sewage from surface sources.

The routine analyses of samples of water from the Southwest land tunnel and the Wilson avenue tunnel has directed attention to abnormal increases in the contamination of the water in these tunnels in passing under the city. As both of these tunnels and their shafts are of concrete construction and drilled through rock the conditions are baffling. Several bacterial surveys of shafts in each system have been made but it has not as yet been possible to locate the source of contamination. The difficulty at the Wilson avenue shaft has been clearing up but at Roseland it has become more serious.

At this station the sampling pump was at first suspected and was removed and carefully sterilized, but the condition still persists. Plans are under way tor an intensive investigation of this tunnel in the vicinity of the pumping station where the condition seems to be localized.


In concluding this report I believe the experiences we have had in Chicago with water supply control work, especially in connection with the contamination of water passing through tunnels, which are of especial interest and importance to water works officials are:

Bacterial analyses are invaluable for supplying information concerning the quality of the water at all important points from the source to the consumer’s tap and should be arranged for in public water supply control work.

  1. The collection of bacterial samples shortly after periods of excels rainfall gives best results in detecting contamination of tunnels due to surface seepage.
  2. Daily bacterial samples over an extended period of time give more reliable information concerning tunnel leakage than a number of samples taken during a single day or part of a day. esi*ecially if the samples of the latter series give negative results.
  3. The collection of extensive meteorological information is an important factor in water supply control work in Chicago where chlorination alone is depended on for disinfection.
  4. The construction of water tunnels under private property and the permitting of private connections to these tunnels are potentially dangerous policies.
  5. Any program which calls for the use of land tunnels for the delivery of a filtered water supply should provide for the complete protection of these tunnels against all possible sources of contamination; and also fer treatment of the water for disinfection immediately before it is pumped to the distribution system.

(Excerpts from paper read before the annual convention of the American Water Works Association.)

No posts to display