w a - a — (0 aaerates Dan ba Ea aL AC SS Ao Since eh db LN SSID Oh STORM SEWER DESIGN HOWARD DANIEL SEVERANCE THESIS FOR THE DEGREE OFC. E, 149) + el a S| < 4 PAs A ab fa Sey Sa SF Goa o Sh AR NRE eee arene OE ek ke eS ee OE OP a | MW Il]! LOD WAN A A Wi (NLO MM co ITN a cor . aw 4 ; & ™ - , YON ‘ : 74 ft ooye f ? ~. a? ~ - ” ° aon, . at | | | PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE Ae d $e 6/07 p:/CIRC/DateDue.indd-p. 1 STORM SEWER DESIGN, THESTS FOR THE DEGREF OF C.F. Howard Daniel Severance, 1913. INTRODUCTION. The subjects of drainage and sewerage are s0 intimately associated as to be almost inseparable. In general,however, it may be stated that while the subject of sewerage refers prin- cipally to the removal of excremental or human waste and otn- er refuse matter common to human habitations, the subject of drainage properly relates to the removal of storm-water from the surface and subsoil. All water given by rain storms may without impropriety be called storm-water. This name,however,is commonly applied to water from rain storms that does not soak away immediately into the ground nor evaporate,but flows away over the surface througn natural chanels or artificial conduits.. Sewers may ve designed to carry storm-water alone,house refuse alone,or both. Those that carry storm-water alone are Called Storm Sewers, those that carry human refuse alone are called Sanitary Sewers and sewers which carry both storm-wat- er and human refuse are called Combined Sewers. A storm-sewer system should be adequate for the prompt removal of the rain fall from the surface during violent storms,including such animal and vegetable refuse from the streets asp will necessar- lly be removed with the storm-—-water. Tne separate system should be built to carry off prompt- ly from houses all sink,laundry and closet wastes without of- fensive odors and without interruption. It should keep itself Page 1. LO37G5 Clean,that is,free from deposits. It should not pollute the soll througn which the pipes pass and it should have an out- let that is without objection. While it is not the purpose of this paper to treat the subject of sanitary sewers or combined sewers,it seems nec- essary to set down some of the advantages and disadvantages of combined systems since the question as to which is best to install confronts the engineer in a large percentage of the cities he is called upon to investigate in reference to the matter of storm sewers. Turing the past twenty-five years great progress has been made in the matter of sanitation,not alone from the standpoint of engineering and the general advancement in sewer construction,but in the public mind generally there has been driven home the fact that sewers are an absolute necessity to the health and comfort of the community. This campaign of education has heen carried to such an extent that it is found necessary vy persons desiring to place subddivi- visions and additions on the market in the shape of town or city lots,to see that their suddivision or addition is proper- ly sewered before offering the same for sale. Yt is indeed difficult to prophesy the good effect of such a spirit even within the next ten years. This condition is mentioned because in the light of wnat is tc follow it will be found that the real necessity for storm sewers does not occur until consider- able time,perhaps years,depending upon the rapidity of city growth,after the sanitary sewers are in place. In general the Page 2, following facts should be considered in case the town or city, or portion thereof under consideration has not any adequate ganitary sewer system. If it has such a system it 1s good policy t» leave it alone and plan a separate storm sewer sys- tem, lst. Sanitary sewage.which constitutes tne dry Weather flow alone in combined sewers is s9 very small in comparison with the storm sewage that in circular sewers,wnich are the most enonomical to build,it forms merely a trickling stream with little velocity over the vottom of tne large sewer re- guired. Tnis condition can be overcome in a small degree by the installation of the egg-shaped type of sewer, wnich can be made of such a snape as to be self-cleaning even with a small amount of flow.Moreover, sanitary sewers are free from sand and other deoris which are inevitably washed into combined sewers during storms and these are especially troublesome in forming deposits. Hence in the separate system generally it 1s easier to keep sewers free from deposits. 2nd. Above the low-water line in combined sewers the ex- tensive interior surfaces of the large sewers require become smeared with filth in time of flood which remains to decay and produce foul gases after tne flood subsides. This coating is not considered vy gome to ve dangerous, However,it 13 a very important factor to be taken into consideration in the design- ing of the comoined sewers. 3rd. On account of the comparatively small size of the sani- Page 5. tary sewers in the separate system it is easier to flusn them 30 as to keep them clean. Automatic flush tanks can be used at small expense to do this very satisfactorily. The combined sys~- tem absolutely precludes the possibility of using the automatic flush tank. 4th. On account of the comparatively small size of the sani- tary sewers of tne separate system the air in them is more fre- quently and completely changed by the daily fluctuation in the depth of sewage and by currents of alr from ordinary ventilation openings. Hence in the separate system ventilation is easier and more perfect. However,in the combined system there is a larger volume of air over the sewage and this dilutes the gases to such an extent that they are claimed to be narmless. 5th. In case the sewage has to be purified the separate system is more economical because only the sanitary sewage need be treated,tne storm sewage veing discharged into nearby water courses. 6th. In small cities and in large portions of large olties the storm water can be usually carried some distance in the eut- ters and then removed by comparatively short lengths of storm sewers laid at shallow depths and discharged into the nearest sultadle natural water courses. In such cases a separate system of sewers will usually cost only a fraction as much as a combin-— ed system. For small towns the great cost of a combined system would either prohibit the construction of sewers entirely or postpone it indefinitely,were it not that a separate system can be bullt so cheaply. On this account alone the introduction of Page 4, . the separate system of sewers has been of incalculable benefit to smaller cities of America. 7th. On account of their relatively small size,sewers of the separate system can be made almost entirely of vitrified sewer pipe,wnich has tne important advantages over brick sew- ers of greater smootnness,of being impervious,of having few joints,and of ease in making the joints practically water tignt. Sth. Often a combination of the two systems can be made to advantage, storm—-water veing admitted to the sewers only in certain portions of the system,such as the business district. 9th. As will be shown in the following,storm sewers are not usually designed to take all of tne 3torm-water in case of extreme storms because in such instances the cost would de pro- hidvitive. In the combined system it would be necessary to make sewers to take such storm-water in addition to carrying the sanitary sewage, flow pecause in the event of a sewer overflowing great damage would be wrought to property and the health of the community endangered. It is to the considerations stated above that the popularity of the separate system and its adoption in many of the smaller cities is due. It 18,moreover,preferable wnere the sewage has to be pumped or purified. Whether or not it should be adopted for larger cities is a matter depending upon local physical condi- tions and economical considerations. In the following pages the subject of storm sewers will ve treated with only an occasional reference to combined or sanitary Page 5. sewers. It is found possible to treat the sudject in general terms only ,owing to the fact that the vasic assumptions to meet dif- ferent local conditions are so greatly varied. It is the earnest desire of the writer to impress the need of careful considera- tions of evonomy throughout the work,from the first assumptions to the final details of construction,for the real worth of an engineer is finally reckoned on the basis of the amount he can save a city or community instead of the amount he expends for it. Unfortunately,nowever,nis pecunary compensation is,in most cases, computed on the latter basis. Economy is not here used to mean the curtailment of the necessary, but the elimination of waste and the unnecessary so far as possidle. The suvject will nere ve treated under the following gen- eral topics; PRELIMINARY INVESTIGATIONS AND TATA. FINAL ARRANGEMENT OF DATA WITH GRNERAL HINTS IN THE MATTER OF MAKING BASIC ASSUMPTIONS. METHOD OF COMPUTING SIZES OF SEWERS. TYTES OF SEWERS AND MATERIALS OF CONSTRUCTION. APPURTHNANCES. The plates and illustrations appended hereto were prepared in the study of a particular problem,altnoughn sone of them are of general application. Their principal value in this work lies in the possibility that they will suggest similar forms and methods, Page 6. PRELIMINARY INVESTIGATIONS AND DATA. Before attempting to formulate any plan for a storm sewer system the entire territory under consideration and such ad- joining areas which may form a part of or be drained by the game system should be carefully studied. A typographical map of the city and adjoining territory should ve made. This map sheuld pe on a scale between 200 and 400 feet to one inch,depending up- on the size of the district under consideration. For undulating ground the contour interval should ve 5 feet:for flat ground 1 foot. A convenient way to proceed is first to make a map of the area on tracing cloth,showing the elevations,contours,and every- thing else that is to be shown on the final map,except special and derived data and the sewers themselves. White background prints can be made from tnese tracings and upon them the dif- ferent classes of data Can be placed and the plan developed rougnily in pencil. In developing plans it is convenient to make figures directly on the map as the work progresses and perhaps several alternate plans may be carried to completion before a final choice is made. Such data as character of so0i1, subsoil, percentage of impervious coverings,street grades witr diréction and rate of fall, character of pavements,number of persons per acre,etc.,shnould ve shown in reference to each block witnin the district. Natural water courses and every possible point of out- let should be located and studied in reference *%o thelr avail- ability in working out a system. A careful study should be made of the general economic conditions in the city,the resources and rate of growth parti- Page 7. Ccularly. This data will furnisn a cue in predicting tne provab-— le future developments,with a fair degree of accuracy if compar- ed with the history of growth in neighboring cities similarly situated. It is a wise precaution to avoid local real estate dealers while drawing conclusions on the sudject of probable future growth. In this connection it may ve said that engineers as a rule, though there are many notadle exceptions,propnesy conditions wnich are unatainable within any reasonable lengtn of time and in consequence assume conditions which fix tne ratios of maximum run-off considerably higher than economic conditions warrant. This 18 probably due to the fear of popular criticism an’ a re- sponse to the clamor “Get the size large enough","“Err on the side of safety",etc. It is preferable to err on the side of safe ty,if at all,bdut not too mich. The probability that a district now untmproved may in fue ture become thickly populated and consequently highly improved also carries with it the certainity that realty values in this district will increase many fold ani can more easily stand the assessment at that time necessary to increase the capacity of the storm system than under present conditions. Besides, this there igs the interest factor which has to be reckoned on the cost of unused capacity which in some instances within the writer's Od - servation would in tne course of twenty years pay for the com- plete installation of a parallel auxiliary sewer system that would more than meet the increased demand of sucn growth. In determining the required capacity of a storm sewer one of Page &. the most important conditions to be considered is the maximum rate of rain fall,that is the maximum rate per hour of pre- cipitation during any given number of minutes. A knowledge of this condition 18 necessary to determine the amount of storm- water reaching the sewer during a storm continuing for 4 per- ioad of time equal to the time of concentration. The records of rain fall are rather incomplete. Records of daily,montnly and yearly rain fall are numerous,but useful as sucn records are for some purposes tney have little value for the design of sewers. The records of storms as generally re- ported,give the total precipitation for the entire storm and provoavly the duration of the storm,but they do not give the Maximum rate of the precipitation. The average rate of preci- pitation, througnout the storm can be obtained ty dividing the total precipitation by the duration of the storm,but this will seldom if ever be the maximum rate o? precipitation,for,as is well known, the greatest intensity of rain fall is attained only during short periods. It is often the case that a rain storm will continue for several hours with a very uneven in- tensity - sometimes a mere drizzle and sometimes a heavy down- pour. Evidently the total precipitation of such a storm will vear no relation to its maximum rate. A storm that will give four inches of rain in twelve hours,may give two inches in two hnours,or provably one and one-nalf inches in thirty minutes. The average precipitation during tne storm would then be at the rate of four-twelfths or one-third inch per hour while the average precipitation during the thirty minutes would be at tne Page 9. a rate of one and one-half divided by one~nalf,or three inches per nour. A locality subject to long-continued drizzling rains may nave a large annual rain fall,while very heavy rains may occur in localities having mucn smaller annual rain fall. It is this maximuiu rate,or rapid downpour during a reasonabiy short per- fod thas wos. oeverely taxes the capacity of a storm sewer.Tne general condition to be considered therefore in designing 4 gewer of this kind is the maximum intensity of the precipita- tion, that is the maximum rate per hour during a period of time sufficient for the water from the most remote parts of the dis- trict to reach the sewer ani flow through the sewer to the point under consideration. This last statement is somewhat general and may be varied by shape of the district under ccnsid-— eration. It is evident that in order to design intelligently a storin sewer the designer should have 1 reasonably accurate record of the rain fall in the locality giving both the rate and the dura- tion of the Various degrees of precipitation for each storm.Sunn records are obtainable by means of self-registering rain gauges in which the continuous amount of rain fall is automatically recorded on a chart moved by clockwork. In 1889 the United States Weather Bureau placed self-registering gauges in the rincipal American cities. Most valuable information relating to the rain fall is given in the Weather Review, the official publication of the United States Weather Bureau. only the recors of self-registering gauges,however,can ve considered as really Page 10. accurate. One important condition indicated by the data ovtained by means of the self-registering rain guage 1s that the maximum precipitation for a short period of time is reasonably uniform throughout the United States. There is,however,some slight dif- ference in the rate of rain fall in different sections. The Southern Coast States,for instance,show a higher rate of maxi- mum rain fall than those of the interior and it is also found that certain mountainous districts also have a much greater rate of maximum rain fall than the more level sections. The rates of rain fall along the Pacific Coast show considerable Variation owing to the proximity of tne Coast Range and the Slerra Nevadas. Generally it is well to take the nearest avail- able data that can be furnished by the United States Weather Bureau and make careful study of the comparative total rain falls in that city and the district under consideration. Rainfall data obtained from the nearest available source snould ve collected and carefully tabdulated,a convenient mode being to set down the rates in vertical columns under their respective duration periods. Page 11. FINAL ARRANGEMFNT OF DATA WITH GUNZRAL HINTS IN THE MAKING OF BASIC ASSOMPTIONS. All data appertaining to the existing conditions as above referred to relating to the sections of the Gity which are now improved and moderately well built up,snoul’ be entered upon the preliminary map. Upon this should also be placed in the un improved districts the assumed conditions,that is,the conditions wnich the storm sewer will be expected to meet within the next forty years. Preliminary sewer location may then be made in rei ink together with tne approximate grade and this sewer location snould extend as far as the requirements of the city to meet the above assumptions,that is to say,when the city is finally built to the limits the Plan outlined upon this sheet should answer the requirements at that time. Drainage areas should ve marked in ink. Adjoining drainage areas may be marked with different colored crayon so as to dis- tinguish them. The areas of these drainage areas should be con- puted and placed in their respective places together with the actual and assumed percentages of impervious surface in each. The density of population indirectly affects the quantity of stormwater flow since it naturally follows that the denser the population the greater percentage of impervious surfaces. Sone authors have suggested a ratio between the population per acre and the percentage of impervious surfaces.Such deductions are gathered from a large numoer of observations and are consequent- ly of a general nature serving as guides rather than laws of procedure. Page 12. The results of such investigations snow that the percent- age of impervious surface is numerically equal to the number of persons per acre in residence districts. This ratio is not true when the density of population exceeds sixty per acre,be- causge beyond this the percentage of impervious surface does not materially increase. This suggests the preblem of future growth upon which there has veen a great amount of discussion oy many cminent authorities. Vain attempts have been made to lay down general laws relating to city growth but it seems that each community has its own mysterious “rule-of—tne-thumo"methnod of growing. There are some very general hints whicn coupled with con- siderable sound judgment will give results accurate enougn for the problem at hand. Study the city carefully with a view of classifying same. For instance,is it a commercial center,a manufacturing center, a resort town,or a suburban residence section. Study the his- tory of its growth and ascertain the agencies whicn have pro- moted it. Make comparison with the history of growth of cities similarly situated and having resources of a similartnature. Just how far into the future it 1s necessary to prophesy to handle the problem with greatest economy 1s a question upon which there is some diversity of opinion,but in the opinion of the writer,as above indicated,forty years is an outside limit even for rapidly growing cities, After determining upon the probable density of population vased upon future growth,percentages of impervious areas can be page 13. computed as above indicated and Placed upon tre sections now sparsely settled and unpaved. The selection of route for main storm sewers 18 a matter requiring careful study in order that the most efficient and consistantly economical location will ve made. In general the sewer should follow the lowest portion of the drainage area under consideration,it veing found economical and sometiues absolutely necessary to cross private property where deep valleys intersect the streets at irregular inter- vals. The cost of sewers is greatly increased when they are built over unstable ground and therefore the character or the soil and subsoil should ve carefully investigated and vad spots avolded if found possible and economical so to do, The particular location in streets,tnat is the particular portion of the street to be occupied by the storm sewer should be such as to avoid damaging sanitary sewers and their laterals, water and gas mains and other permanent structures,due precau- tion being taken to avoid excessive cuts. The writer has in some instances found that the most economical location was the center line of the sidewalk in residence sections. At others as near the center line of the street as possible,keeping above anid parallel to the sanitaryssewer,at one side or the other of same. When the location is made in sidewalks less depth of earth ani other material over the sewer 1s necessary than when the loca- tion is made in the roadway portion or the street. Plate I shows a location of main sewer. The final location to meet the present needs is shown in full line black ink. Proposed exten- Page 14, sians to meet future reguirenents are shown dotted. The alter- nate location,abandoned because of economical conditions,1is shown in red ink. This is mentioned merely to indicate that in the final location of storin sewers economical considerations are really paramount. Some working equation for rate of rain fall or curve snow- ing the relation vetween tha rate of rain fall and duration of storm is of prime importance to the designer. A general form of equation widely used and nignly recom- mended vy vest authorities is: a t+ovn r=- in which r is the rate of rain fall in inches per hnour,t is the duration of storm in minutes and a and bd are constants. Tne selection of values for a and b may be made vy plottirz on cross-section paper,tne records of storms of various durations taken from data obtained and tabulated as indicated in the discussion of “Preliminary Data",and selecting typical maximum values of rate "r*“" at various correspond‘rg values of duration *t". Then substitute these corresponding values of r and t in the above equation as constants. The values of a and b,treated as unknown quantities can then be calculated rrom any two such equations, considered simultaneously ,as follows: Assuming that r, ,t,and r,,t- are corresponding values for r and t, then r= a and rc a t, + b t.+ b from which a-=frt uit t.- I. t, ) ‘ ‘ YU Te and v Ye ter rt r,- Ye Tne general relation between a and b 13 approximately such that b = 13%. The rollowing values for a and b are guggested vy Prof. A.N.Talbat: yor rare storms a = 360, c= 30: For occasional storms a = 200, v= 20: For ordinary maximum storms, a = 105, v= 15. After determining the proper Value Of tudse Countaiuts a working equation can be made and a curve,similar to Plage II, drawn for convenient reference, The ordinates give the rate of rain fall in incnes per hour corresponding to the duration of storm shown along the avscissa:. This curve was designed for a particular Local ity in which there may be storms whose rates will exceed those shown vy this curve,but the probavility of their occurrence is estimated as once in ten years. The rain fall in cubic feet per second per acre corresponds almost exactly,or close enough for all practical purposes, to the rate of rain fall in inches per hour. It is safe to assume that total precipitation in cubic feet per second per acre equals to the rate correspondingthe duration of storm. The chart shown on Plate II] 1s derived from the preceding curve and the equation therefor. The ordinates OY on this sheet represent the area to be Grained in acres,the avscisse. OX the total amount of precipita- Page 16, , tion in cubic feet per.second. The lines 0 = 5, 0 - 10, 0 ~ 15, etc.are tne loci pf points whose ordinates are,area and total precipitation, corresponding to the times of concentration,five mimtes,ten minutes,fifteen minutes,etc. ,respectively. To find the total precipitation on a given area,say four acres,corres- ponding to the time of concentration of say twenty wintites,pro- ceed horizontally from the ordinate fcur to its intersection with the vector 0 —~ 20,thence downward vertically to the ab- scissa 0 ~ X,which will give the total precipitation in cubic feet per second as twelve. This chart is a specially derived one and 1s found to ve very convenient for checking off results of analytical computation. Page 17. METHOTS OF DETERMINING SIZE OF SEWER. When rain begins to fall upon an area drained by a storm sewer the water falling in the immediate neighborhood of the outlet at once enters the sewer and begins to be discharged. As time passes and the rain continues water arrives at the outlet from tne more remote portions of the drainage area and the discharge at the outlet increases quite rapidly until wat- er is being discharged from all portions of the drainage area at the same time. After that any further increase is slow,being due only to a percentage of run-off slowly increasing as the saturation of the soil becomes more complete. The timed concentration is the longest time required for water from the remotest portions of the drainage area under caon- sideration to reach the outlet of that portion. For general considerations the storm causing the greatest rate of discharge in a storm sewer will ve the maximum storm lasting a length of time equal to time of concentration. If a time less than this ve taken the water will not ve discharged at the outlet from all the points of the drainage area at once and that very near the outlet will have a chance to run away before that from the remotest point arrives. On the other hand,if a time be taken longer than the time of concentration the heaviest rate of the Maximum storm lasting this long will be less than the rate of the maximum storm lasting a length of time just equal to the time of concentration,and since the storm i383 lighter the flow will be lignter. Not all the water falling ina drainage area will be carr® 4 Page 18. away in the sewer. During and after the storm some of the water evaporates into the air and some is absorbed into the soll. Some also collects on the surface to flow on into tre sewer af- ter the storm 1s ended. The engineer jJetermines the percentage of rain flowing off in the sewer by estimating the maximum percentage of run-off of the drainage ares. The general method of calculating the amount of storm sewage for any particular drainage area is as follows:- (a). Calculate the time of concentration or longest time of flow to the point for which the size of sewer is to be de- termined. (bd). Calculate the maximum rain-fall corresponding to the time of concentration, (c). Calculate the percentage of impervious and pervicus area on the water shed drained oy the sewer. (d). Using the percentage of impervious and pervicus areas obtained as above,calculate the maximum percentage of run-off or the percentage of fhe rate of maximum rain fall whicn will ve running off in the sewer under design at the end of the time of concentration. (e). Calculate the total maximum rate of flow of storm sewage by multiplying together the drainage area,the maximum rate of rain fall corresponding to the time of concentration, and the maximum percentage of run-off. (a). The time of concentration may generally be divided as follows: l. The time required vy the water from the roofs and yards Page 19. to reach the pavements or the gutters. 2. The time required to run along the gutters to the in- let ani the sewer,and 3. The longest time required for the Water to flow through a line of sewers to the point for which the size is to ve calcu- lated. l. The time required for water to reach the gutters may usually be estimated vetween five and ten minutes,depending up- on the slopes and surfaces and also upon the density of the pop- ulation. In thickly built up sections the rate of run-off from yards,roofs,etc.,may be taken at the lower rate while sparsely settled porticns of the city,with rlat surface slopes may ve estimated at ten minutes oreven higner. 2. Time required for water to flow along the streets and gutters to the sewer may be calculated from the formula v-cv8 in which v is the velocity in feet per second,c is a constant depending upon the character of the surface and s 18 equal to the fall in feet per hundred feet. Plate IV snows the curves for different types of 3urfaces and is useful in determining velocity for various percentages of siope. Tnese curves will also be found useful in odtaining velocities of flow over outsite and undivided areas whicn are tributary to the syste. 5. Tne time reguired for water to flow through the sewers to the point of consideration sghovld ve computeil voy means of Kutter's formula,or by reference to the curves of velocity and discharge on Plate VI appended hereto. For exict determin- inzgs,however, the velccities.flowinz full,are a little too low Page 20. a8 will be seen vy reference to Plat2 VII showing the relative values of velocliy uut discharge for aifferent deptns of flow. tn this plate the three curves represent,respectively, the re- lative values of the areas of wetted cross-section.of the ve- locity and of tne discharge for different deptns of flow, tne value for the full depth of flow being taken a3 unity in eacn case. For any deptn of flow the value given oy the area curve multiplied vy the vaiuegiven vy the velocity curve will equal the Value given oy tne discnarge curve. The relative value given ‘by the curves of the dilagram are for a uniform Slope provitei tney all relate to the Same siope. The curves of velocity and discharge are necessarily only approximately exact,but they are sufficiently accurate for ail ordinary computations. All tne curves on the diagrauw relate to circular sewers. Kutter's formula is useG in all the computations in this paper and coin- putations are usually made on the basis of the sewer flowing full. However,sewers rarely discharge at their full capacity and in practise should not ve designed so that they will do so, The velocities that are shown in the diagram will not be the usual velocities,but will bear a relation to the actual velo- cities that can be ascertained according to the deptn of flow. This curve is particularly useful in determining the time of concentration because it shows that the values given by Kutter's formula when the sewer runs full really are not maximum and nence the time of concentration would be too long if such formula were used without regard to the maximum velocity possible to attain in a circular sewer, Page 21. (bd). In calculating the maximum rate of rain fall corres- ponding to the time of concentration substitute the time of concentration t, as obtained by the above method in the equa- tion for maximum rate of rain fall or use the element t as an argument in using the curve plotted from the same equation, which will give directly the value of rate of rain fall corres- ponding to that particular time of coneentration. A chart simi- lar to that shown on Plate III may be used to advantage at this point to obtain the total precipitation in cubic feet per second on the area under consideration corresponding to the computed time of concentration. Select the nearest time vector or interpolate if necessary and proceed as explained in the pre- ceding section. (c). The percentage of impervious area may be calculated in the following manner: Take a typical unit of area,usually an average vlock naving different percentages of imperviousness as follows:- Roof area. From the average size of buildings and the aver- age number of vuildings per block actually in existence or as- sumed to meet the conditions of further growth,wnich will be comnected with the sewer,or with the gutters, Calculate the total rovf area in the block. Take this at its full value if the roofs are connected with the sewers, but only ninety per cent if the roofs are connected with the gutters. } Pirst-class pavements. Calculate the total area per vdlocx of bricks,asphalt and stone bdiock with tight joints and take eighty per cent of this area. Page 22. Second-Class pavements. Calculate the total average area per vlock and take sixty per cent. Third-class pavements. Calculate the total average area per bdilock of good macadam and similar pavement and take forty per cent. Hard earth roads. Calculate the total average area per block of the traveled hard earth surface and take twenty-five per cent. Sidewalks. Calculate the several total average areas per block of first,second and tnird class sidewalks corresponding to the classes of pavement as above and take the same percent- ages as for the corresponding classes of pavements,namely eighty, sixty and forty per cent for first,second and third class side- walks respectively. But if the pavements are separated from the gutters by side parkings,as in some city residence districts, take only one-half of the above percentages. Finally ad1 together all the rated average areas per block odtained as above explained and divide the sum by the total area Of the typical block. The quotient will be the percentage of impervious area. The percentage of pervious area is obtained vy subtracting the percentage of impervious area from one hun- dred per cent. Another method which gives excellent results and is almost indispensable,in case the approximated future conditions are considerable in advance of the present,is to use the popu- lation per acre,actual or estimated,as a basis for estimating the percentage of impervious surface. The rule that the percentage of impervious surface is equal, Page 23. numerically,to the number of persons per acre will give results consistent with the accuracy of other assumptions necessary in the general problem. (a). Calculation of maximum percentage of run-off. AS be~ fore stated,not all of the water or rain falling on the imper- vious area of a water shed will run off during the storm. More or less amounts are evaporated or absorbed at once for no sur- faces are absolutely impervious. A large amount goes to fill up small depressions in the surfaces,a still larger amount accum- lates on the smrfaces of the water-shed making its way toward the sewer. The amount so accumulated and its rate of movement increase$s,as the storm continues at the same rate,until final- ly an equilibrium of flow is established and the rate of run- off from the impervious area becomes practically one hundred per cent of the rain fall;thnat is,the snorter the storm the less per centage of run-off,an’ hence the sewer water-sheds having smallest time of concentration are likely to have the smallest per centage of maximum run-off. The maximum downpours which de- termine the size of the sewers,are often preceded by lighter downpours which saturate and partially flood the water shed, hence it will probavdly never be reliable to assume less than seventy-five per cent as the percentage of the maximum run-orf from the impervious area of a sewer water shed even for short times of concentration and comparatively little damage from over-charged sewers. With longer times of concentration, say forty-five minutes or over,ninety-five per cent of maximum run- off from the impervious areas should be aswumed. In the case of Page 24. long-continued storms pervious areas become gradually saturat- ed until some run-off occurs from it also. In case of storms lasting for several hours,such as caused the great floods in the rivers, this percentage of maximum run-off may de quite high,but for sewers the time of concentration and hence the duration of the maximum dowmpour are comparatively short, rare- ly longer than one hour. The above statements are necessarily quite general in thet nature,since it is impossible to lay down any fixed rule for ov- taining the percentage of run-off. Careful study of local condi- tions and close observation of conditions of streets and exist- ing sewers during storms are necessary links in the data chain. Localities which are subject to long continued rains, that is,such as will completely saturate the grouni,will produce a higher percentage of run-off than the contrary. The texture of soil and subsoil and general slope of the ground still further varies this percentage value. The writer has prepared a chart shown on Plate V,which serves to illustrate the relation of percentage of maximum run- off to duration ofstorm for a few different typical conditions. The ordinates show the percentage of run-off and the ab- scissae the duration of storm corresponding to the computed time of Calculation. ) The upper Curve marked "Impervious surfaces" represents as near the actual values as can be obtained from considerable ob- servation and dest authorities. Page 25. Tne three other curves are typical and may or may not have practical application to a particular problem, However, they illustrate the generally accepted proposition that the percentage of maximum run-off from pervious surfaces bears a direct ratio to the time of concentration. It must be vorne in mind that the percentages given by this chart are not the percentages of run-off for the entire surface out percentage of the percentages of the various Cclass- es of surface. For instance,if there ve forty per cent of the surface impervious and the remaining sixty per cent pervicus, (hard soil,steep slopes, and the time of concentration twenty minutes, the total percentage of run-off would be ~86 X 40 + .13 X .60 = 42.2% of total precipitation running off during the time under con- sideration. (e). The total effluent or maximum rate of flow of storm sewage at the point on the sewer line under consideration is the continued product of;(1) the area in acres,of the district drained vy the sewer at that point,(2) the maximum rate of rain fall,in inches per hour,corresponding to the time reguir- ead for the water to reach the point under consideration from the most remote portions of the district,and (3) the maximum percentage of run-off. The proper size of sewer to carry the quantity of water as apove calculated at the grade which the sewer will nave leading away from the point under consideration may be comput- ed from Kutter's formula. Page 26. Plate VI shows a graphic relation between the size,grade and discharge of circular sewers running full. This was com- piled vy using Kutter's formula and assuming the coefficient of roughness NW = .015 which 1s a safe value for vrick or con- crete construction. The values given may ve reduced s© as to give values of velocity and discharge for NW =.013 vy adding nineteen per cent t the values taken from the curves. Page 27. TYPES OF SEWERS AND MATFRIALS OF CONSTRUCTION. Plate VIII shows three common types of sewer-section,name- ly,the circular,ege-shnaped and the U-shaped. Upon the same plate will be found the relative values for area and velocity of discharge. The circular type of sewer is an old favorite vecause of its simplicity and because of the economy of its construction. The egg-snaped type is used for combined sewers in which there is a small dry-weather flow,a smaller amount of flow giving a higher velocity than in a circular sewer. Tne U-shaped sewer is recommended for size not over forty-two inches that is forty-two inches in greatest dimension. As seen from the characteristics of this type it has a higher velocity and discharge value tnan a similar size of circular or egg-shaped sewer. It 18 a good type to use where the sewer is placed under the sidewalks,with small head roon. Special cross-sections are sometimes found necessary for large sewers,of which there are a great many varieties used in the various cities of the United States and of Furope. | In general the selection of type of section depends upon the grade,location and quality of sewage to be carried. The writer has found it economical to select and specify several different types of construction which he considered equal in efficiency and to use the type found to be lowest in cost of installation. This method has a further afivantage of stimulating competition. The engineer may find that this can- not be done in some cases where a particular type may prove the Page 28. most economical even though the first cost is somewhat higner. Materials of construction for storm sewers are,vitrified pipe, stone, brick,and concrete. Sewers twenty-four inches in diameter and under are usual- ly built of vitrified sewer pipe. This material has a smaller coefficient of roughness than any other. Care should be taken, hnowever,in the use of vitrified pipe to see that provision is made for the removal of sand and other damaging detritus,es- pecially where the sewers discharge under high velocities, The writer has noted several sewers laid of this material and nas found in some instances that the sewers have veen cut nearly through in the snort time of five years. Stone was formerly used in the construction of sewers to considerable extent,especially in the large sewers. On account of its nign coefficient of roughness and the high cost of good stone masonry this material is rarely used in modern times. Brick is a favorite material for sewers too large to be made of pipe. The dividing line is usually drawn at thirty to thirty-six inches. Brick present many advantages for sewer wx, including their moderate cost and their durability and their small size and regular shape,wnhich enable them to be handled readily and used in building sewers of any desired cross-section with comparatively smooth and true interior surfaces. Sewer brick,as those suitable for sewer construction are commonly called,snould be harder burned than ordinary building brick to enable them to stand the wear from the flow of sewage and to insure against disintegration. They need not,however,be Page 29. as hard as No.l] paving brick and hence constitute an interme- diate grade between building brick and pavers. Sewer brick should ve uniform in size and of reguilar,true shape so as to prevent their being laid with thin Joints to form smootn, true surfaces. Great care is necessary in the construction of sewers with tnis material to make the joints water tight and it is be- cause of this difficulty that concrete is becoming a favorite over vrick construction. Concrete,of late years,nas been frequently employed in pre- ference to brick or other kinds of masonry. It has the following advantages: The cost is usually less than the cost of vrick masonry. Tne concrete exactly fits the irregularities of the exca- vation and hence gives better foundations. There are no joints as in brick work t> be made water tight though on the other hand it is not easy to make the body of the concrete entirely impervious to seepage. Concrete can be readily moulded into any desired shape of sewer. Concrete can be made by comparatively unskilled workmen if a skilled foreman is employed. The concrete pipe,similar to that snown on Figute 2,Plate Ix is rapidly coming into favor,especially that of the reinforced type. Tnis pipe is made in sections usually three to four feet in length and of the diameter required,and reinforced with longi- tudinal and transverse bars. It is claimed that this method of making concrete sewers is superior to that of the monolithic type Page 30. vecause there is a better orportunity afforded for thorough in- spection of each unit before it is laid in the sewer. Also it is possible to make the concrete considerably denser by the use of water tight moulds. It is further claimed as an advantage for this type of construction that the shrinkage which takes place while the concrete is settling is entirely accomplished vefore the pipe is laid in the line of the sewer. This avoids the cracks which are numerous in monolithic construction which is attributed to the above-mentioned cause. Some authorities on sewer construction recommend paving the invert of concrete sewers with brick as it is claimed that brick is less subject to erosion than plain concrete. In the lignt of evidence furnished by city engineers from a large nun- ver of the cities of tne United States,tnis conclusion cannot be based upon observation of well-made concrete sewers. It appears from such data that concgete,when the ingredients are properly proportioned,is properly mixed and handled,and carefully cured, has proven equal to and in some cases superior to the more ex- pensive brick construction. The writer therefore has no heSitancy in recommending the use of concrete wherever its use is economical from the stand- point of first cost provided the materials obtainable are first- Class and provided further that the soil or sewage does not give strong alkali or acid reactions. plate IX shows four ordinary types of concrete sewer. Ficure 1 illustrates a type used in soft,wet greund. Figure 2 illustrates a type of reinforced concrete pipe sewer. Figure 3 the U-shaped concrete sewer,and Figure 4% a common type of reinforced mono- Page 31. lithic sewer for good 8011 conditions. Page 32. APPURTENANCES. Street inlets and catch basins are the means of admitting water from the street surfaces into the sewer. The storm inlet implies merely a branch sewer with a grated opening on the curb, while the catch basin has in addition a silt well or basin to catch sand,silt and other detritus preventing their entering the sewer. In combined sewers both inlets and catch basins snould be trapped. Yor storm sewers generally the catch basin is an economic failure and more or less of an offense to modern sanitary ideas. Theoretically they are supposed to eliminate all foreign sub- stances and keep tnem from entering the main sewers,and if they can be Cleaned after each storm they are not odjectionable. The fact remains,however,tnhat they 40 not eliminate all the sand, silt and other detritus because they are not cleaned after each storm. The attention of city officials generally is called to so many things which are above the surface that need doing that they find very little time tn attend to such things as catch basins. If, nowever, they were attended to religiously the cost of so doing would ve too great to fustify their possible advantage over the direct inlet. There are two conditions under which the installa- tion of the catch vasin for storm sewers are justifiavdle,namely, " when the velocities in the main sewers or in portions thereof be- low the inlet or catch basin are very low or extremely hign,which conditions would allow,in the first instance,deposits to form in the sewers and wnen the velocities are extremely hign the pre- sence of sand and other detritus of a grity nature has a tendency Page 33... to erode the pipes. The engineer,in designing the storm sewer systems should take into consideration the character of the detritus which is apt to be wasned into the sewers and if it igs of such nature as is not apt to form deposits or erode tne pipes of the sewers then the catch basin should not be install- ed. There are many types of gratings to be used in connection with inlets and catch basins now in the market. Mm selecting a type the following requirements snould be fulfilled. The grating should be such as to eliminate devris such as large sticks and voards. They should be so arranged as to offerno. ovstruction to traffic and the opening should be large enough to accommo- date all water that will arrive at that point during storms. The combination of the horizontal bars placed in the gutters at right angles to the curd with the horizontal bars across the opening in the curd itself makes a very efficient arrangement for grating.Simple types of inlet and catch basin are illustrated on Plate X,figures 5 and 64. Mannoles should be placed at intervals along the line of the storm sewer,particularly at points where grade changes and also where the alignment changes. These manholes offer the op- portunity of inspection and when placed at the street intersec- tions permit the vest arrangement for connections to catch vdasins and street inlets. A few types of manholes are shown on Plate X, Figures 1 to 4% inclusive. Manholes mst be built large anough at the bottom and for a couple of feet above the top of the sewer to permit a man to work comfortably. Four feet in diameter Pave 34, is a satisfactory size. Sometimes mannoles are eliptical at the bottom with the long axis lengthwise of the sewer. This form is a little more expensive vecause of being more diffi- cult to build. Above the top of the sewer the manhole should ‘be gradually drawn to a diameter of about two feet. The cover Castings may be of any manufacturers design satisfactory to the engineer,weigning at least three hundred and seventy-five pounds. For sanitary sewer manhole the covers are usually per- forated with one-inch holes to permit ventilation and below it there is hung a heavy castiron dust pan to catch any dirt en- tering through the perforations. For strictly storm sewer con- struction,however,the perforations in the cover and the dust pans may be omitted. Manholes are usually built of brick, in fact brick is the most economical material to use in the con- struction of same. Some conditions,however,wnere special de- signs are Called for require that concrete be used. In working out the detailed drawings for manholes,catch basins and other appurtenances to the storm sewer it is well to make the plans as explicit and perfect as is possible to make them because it is found easier to exact perfect work of the builders under perfect plans,than with rough, imperfect ones. Manholes should be placed along the sewers at intervals of not over fcur hundred feet and at all points where there are ab- rupt changes in alignment or grade. Very large sewers 4° not re- quire manholes at such frequent intervals as smaller sizes. Page 35. PLATE VY. RUN-OFF DIAGRA PLATE lf. ee PLATE VII. PLATE VIII. PLATE IV. PLATE V. RUN-OFF DIAGRAw PLATE IX. * hee ~/b & sav we de ye oy gg ys vl eet Ar } Lo a“ y o “vs se ae s ee ae ? tA Lue? LEI MACHA Figs Be ore esr cerrrerce ere 4 “rt , ? , i ’ i : : ha t i 1 ‘ te L J : . pobnne Bere eees wee. a “rt PLATE X. Fig. 1. Fig. 2. Fig. 3. LNLET Fig. 4. Fig. 5. Fig. 6. + en, * wes. ponent ao” one wee MICHIGAN STATE UN iii 50026 ii me Aran ENA AAT NASER OER ATA SEA ATRIA AA ATA nad ATA AAT TAROT Re rene anh a