71 11,915 - MC GUIRE, Michael Edwin, 1938AN APPROXIMATION OF MICHIGAN'S SPATIAL WATER NEEDS. Michigan State University, Ph.D., 1970 Geography U niversity Microfilms, A XEROX C o m p an y , A nn A rbor. M ichigan AN APPROXIMATION OF MICHIGAN'S SPATIAL WATER NEEDS By Michael Edwin Me Guire A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Resource Development 1970 ABSTRACT AN APPROXIMATION OF MICHIGAN'S SPATIAL WATER NEEDS By Michael Edwin Me Guire Water has been in the past, and will continue to be in the future, of vital importance to Michigan's continued growth and development. Demands on the state's water re­ source base are increasing, however, and if all of the var­ ious uses of water are to be satisfied forecasts of future use will have to be made. The purpose of this study, there­ fore, was to examine the factors which have been instrumen­ tal in affecting the present spatial pattern of water use in Michigan, to consider changes in these factors over time, and to estimate what the future demands for water in Michigan may be. Water withdrawals in any area have a certain struc­ ture, the principle withdrawals in Michigan being made for domestic, municipal, and commercial uses, industrial uses, and for agriculture. Estimates of future spatial water d e ­ mands were made by applying unit withdrawal factors to fore­ casted magnitudes of the withdrawal u n i t s . Withdrawals for domestic, municipal, and commercial purposes are expected to increase at a rate greater than the corresponding growth of population, a result of improved socio-economic status. These estimates were made by m ulti­ plying a per capita withdrawal rate times the projected population. The results indicate that domestic, municipal, and commercial withdrawals were approximately 162 billion g al­ lons per year in 1940 and had grown to over 250 billion gal­ lons by 1960. 13y the year 2000 annual withdrawals are ex­ pected to be almost 500 billion gallons. Withdrawals for manufacturing purposes were estimated by multiplying a per employee withdrawal rate times the num­ ber of manufacturing employees expected in the state. Manu­ facturing withdrawals are believed to have been approximately 275 billion gallons per year in 1940. With the rapid growth of manufacturing this increased to about 775 billion gallons in 1960. By 2000 withdrawals for manufacturing should be over 1,000 billion gallons per year. Domestic, municipal, and commercial withdrawals and withdrawals for manufacturing are highly concentrated in the southern third of the state; 15 counties account for over 80 per cent of these withdrawals. In addition, manufacturing withdrawals were concentrated in four manufacturing types. Approximately one-half of one gallon of water is re­ quired to produce one kilowatt minute of electricity. In 1940 this amounted to slightly over 330 billion gallons per year. With the expected growth of population and economic activity this may increase to over 6,500 billion gallons per year by 2000. Most of the withdrawals will be concentrated along the Great bakes because of the tremendous quantities of water available. Rapid changes in Michigan's agriculture has precluded specific estimates of withdrawals for agricultural purposes. However, the expansion of irrigated acreage and technological developments in livestock production are expected to affect significant withdrawal increases. These should be concen­ trated in the southeastern and southwestern parts of the state, the major areas of irrigation at the present time. ACKNOWLEDGMENTS The participants in this thesis are legion and it would certainly be some form of plagiarism not to give full recogni­ tion to each. Unfortunately space does not permit this oppor­ tunity and a less ambitious attempt must be made. Greatest help was received from members of my family who, through both encouragement and freedom from family re­ sponsibilities, provided an environment conducive to study. Perhaps over the years the debt can be repaid. Faculty help from the Department of Resource Develop­ ment has guided me around numerous pitfalls and renewed my confidence when this was sorely lacking. cially Appreciation espe­ is due to Drs. Barlowe and Steinmueller whose doors were always open. Fellow graduate students in the department combined to establish a situation in which academic standards were held high. without the competition offered by these men the com­ pletion of this thesis would have been in doubt. Finally, gratitude is expressed to the Michigan Water Resources Commission whose financial assistance permitted un­ divided attention to the thesis for a year. ii TABLE OF CONTENTS LIST OF TABLES .................................. . . . . . V LIST OF F I G U R E S ............................................. V i i LIST OF A P P E N D I C E S .......................................... Viii Chapter I. BACKGROUND TO THE S T U D Y ........................ 1 Purpose Conceptual Framework Objectives Scope Introduction to Methodology Assumptions II. FACTORS CONTRIBUTING TO DOMESTIC, MUNICIPAL AND COMMERCIAL WATER WITHDRAWALS ............. 22 Introduction Composition of Withdrawals Elements of Domestic Withdrawal Increases III. GENERAL POPULATION CHARACTERISTICS AFFECTING DOMESTIC, MUNICIPAL, AND COMMERCIAL WATER WITHDRAWALS IN MICHIGAN .............. 35 Introduction Population Growth Changes in Population Characteristics Population Projections IV. DETERMINATION OF DOMESTIC, MUNICIPAL, AND COMMERCIAL WATER WITHDRAWALS .............. Introduction Possible Approaches Data Availability and Deficiencies Examination of Possible Approaches iii 51 Chapter V. FACTORS CONTRIBUTING TO MANUFACTURING WATER W I T H D R A W A L S .................................. 77 Introduction Nature of Water Use in Manufacturing Manufacturing Activity in Michigan VI. DETERMINATION OF MANUFACTURING WATER W I T H D R A W A L S ................................. 110 Estimates of Manufacturing Employment Water Withdrawal Estimates VII. WATER WITHDRAWALS FOR POWER GENERATION .......... 148 Importance Nature of Withdrawals Ouantity Withdrawn Location of Withdrawals VIII. AGRICULTURAL WATER WITHDRAWALS ................... 154 Importance Land Use in Michigan Water Use in Agriculture IX. SUMMARY t FINDINGS , AND R E C O M M E N D A T I O N S ........... 211 Summary Findings Implications and Recommendations BIBLIOGRAPHY ............................................... 228 A P P E N D I C E S ................................................... 236 iv LIST OF TABLES Table Page 1. Michigan Surface Water Resources ................. 2. Price Elasticity of Residential Water Demand 3. Population Growth: US, ENC, M i c h i g a n ............ 37 4. Regional Population Concentration .............. 45 5. Michigan Urban and Rural Population ............ 49 6. Regional Service from Municipal Systems 7. Regional Non-Industrial Withdrawal Coefficients .................................... 76 8. Employment Trends in Michigan ................... 88 9. Total State Employment, Projections to 2020 ........................................... 117 10. Total State Employment, 1964-1968 . . . . . . . 118 11. Regional Manufacturing Employment .............. 132 12. Error R e d u c t i o n ...................................... 141 13. Water Withdrawal Coefficients 14. Number of farms; State Total and Regional Distribution ........................... . . . . . . . . . . . . . . . 2 31 70 145 175 15. Loss of Farms in M i c h i g a n .......................... 177 16. Land in Farms: State Total and Regional Distribution .................................... 178 17. Average Farm Size in M i c h i g a n ..................... 179 18. Cropland Harvested in Michigan, ............ 180 19. Irrigated Acreage: State Total and Regional Distribution .......................... v 183 1964 Taole 20. Page IJumbor of Livestock: State Total and Regional Distribution ......................... 205 21. Livestock Specialization in Michigan ......... 206 22. Summary of Michigan's Water Withdrawals . . . . 214 23. Domestic, Municipal, and Commercial W i t h d r a w a l s ...................................... 217 24. Composition of Manufacturing Withdrawals by T y p e ........................................... 219 25. Regional Composition of Manufacturing W i t h d r a w a l s ...................................... 220 vi LIST OF FIGURES Figure Page 1. Planning R e g i o n s ................................ 14 2. Normal Potential Evapotranspiration ............ 28 3. Normal D e f i c i t ................................... 29 4. Price-Deraand for Domestic Water: Western Cities .................................. 32 Michigan Demographic Rates: Live Births, Deaths, Natural Increase 1900*1964 ........... 40 6. Population L o s s ................................... 43 7. Effective Buying Income Per Household, 1968 . . 47 8. Relationship Between Population Served and Total Withdrawals ......................... 66 Total Employment and Per Cent Manufacturing . . 90 5. 9. 10. Total Employment ................................ 11. Manufacturing Employment 12. Installed Generating Capacity 13. Power Plant Location, 1970 14. Length of Growing S e a s o n ..........................161 15. Major Irrigated Crops, 1964 ..................... 195 16. Irrigated Acreage, 1964 .................. 196 17. Major Cattle, Hogs, and Sheep Producing Counties ............................. 209 vii ....................... ................ . 120 128 149 153 LIST OF APPENDICES Appendix Table Page 1. Sources of Michigan Population Growth .......... 236 2. County Population Concentration ................. 237 3. Comparison of Population Estimates ............ 238 4. Non-Industrial Withdrawals, United States . . . 239 5. Independent Variables for Regression M o d e l ............. 239 Per Capita Non-Industrial Withdrawals, 1965 ............................................. 241 Per Cent Served by Municipal Systems, M.D,P.H. Regions ............................... 242 8. Increase in Withdrawal Rates 244 9. Manufacturing Composition, Michigan . . . . . . 244 10. Regional Manufacturing Employment .............. 245 11. Structure of Regional Manufacturing E m p l o y m e n t ......................................... 246 12. Spatial Composition of Regional Manufacturing Employment ...................... 249 Manufacturing Employment as a Proportion of Total Employment ........................... 256 Regression Coefficients for Lower Peninsula Regions ............................. 256 15. Regression Coefficients for Counties .......... 257 16. Derivation of State SIC Projections ............ 259 17. Regression Coefficients for State SIC E m p l o y m e n t ...................................... 259. 6. 7. 13. 14. viii ................... Appendix Table 18. Page Water Recirculation .............................. 260 Appendix Figure 1. Michigan Department of Public Health Planning Regions ................................. ix 243 CHAPTER I BACKGROUND TO THE STUDY Purpose That water is of inestimatable value to an area is seldom disputed. Whatever the nature of the area, unless it is completely unpopulated, water is of major importance. It is not necessary to expand the discussion of this point other than to say that in virtually all aspects of water use the demand is increasing in almost every area of the nation. Pop­ ulation growth and the expansion of economic activity dictate that increasing amounts of water will be utilised in the fu­ ture. Water use in Michigan is particularly important. Mich­ igan ranked seventh in population in 1960, and in value added by manufacturing in 1963 it ranked sixth. Thus from sheer size, vast amounts of water are required and more will un­ doubtedly be required in the future as population and eco­ nomic growth continue. In Michigan water quality has not been a major prob­ lem. With location on four of the Great Lakes, with a large amount of inland water, and with abundant groundwater re­ sources, the state has not felt the pressure of the demand 1 2 for water impinging on a limited supply as have other states, une measure of Michigan's surface water resources is indi­ cated in Table 1. Table 1 Michigan Surface Water Resources* Source Area Great Lakes 24,688,000 Acres Inland Lakes 841,000 Acres Inland Streams 36,350 Miles *Clifford Humphrys, Michigan Lakes and Ponds (East Lansing: Agricultural Experiment Station, Michigan State Uni­ versity, 1965), quoted in Raleigh Barlowe, Implications of Land and Water Use Developments in Michigan tor Future public Water Resource Policy (East Lansing: Department off Resource Development, Michigan State University, 1966), p. 1. If all of the various uses of water in the state are to be satisfied in the future expectations of future uses will have to be made. Although adequate water resources to satisfy present uses are available there is no assurance that they will be adequate in the future. Water policy made today will affect the adequacy of water resources tommorow. These policy decisions can be made with some degree of certainty if some idea of future requirements can be obtained. The developments which influence the amount of water utilized have not in the past been equally important in all areas of the state. Some sections of the state, notably the northern parts, are probably using less water in 1970 than in 3 19 50. On the otlier hand, most of the overall increase in population and economic activity, and as a result increases in water use, are taking place in 15 or 20 of the southern counties. If the supply and demand situation for water is to re­ main adequate information must be attained on the structure and amount of water use throughout the state. Consequently, the purpose of this study is to examine the factors which have been instrumental in influencing the present spatial pat­ terns of water use in Michigan, and to consider the changes which may be expected in the future. The focus of this ef­ fort will be on the non-marginal changes which have taken place over the years in the elements of water use in order to indicate the probable developments in the next few decades. Estimating water use for the state as a whole is a very complex matter and involves the use of more resources than were available for this project. Nevertheless, in order to give a better indication of what may be expected in the future patterns of water use, tentative estimates of selected uses will be made for I960 and 2000. Although these esti­ mates may, in themselves, be important to personnel concerned with planning for Michigan's future water requirements, it is believed that the value of these use estimates will not be so much with the amount of use predicted, but with the identifi­ cation of spatial trends. Water policy established which is based upon an appreciation of the spatial aspects of water 4 use will be a most; important addition to the planning effort. Conceptual Framework A prerequisite for the initiation of any research is the establishment of parameters within which the research is to be carried out. Research which does not contain this essential element will very likely bewilder the reader with conflicting terms and with an unclear idea of the purpose and direction of the work. For the researcher, too, the early setting of the framework will better insure that the research plan is followed and that the purpose and objectives of the study are completed. The major concepts which must be established at this point are the nature of water use and the types of water use with which this study will be concerned. The Nature of Water Use Water may be used but it is never destroyed in the sense that it is diminished in physical quantity. The form of the water may change, causing it to be less available for use, but the potential amount of water remains unchanged. This is exemplified by the concept of the hydrologic cycle whereby the world's potential supply of moisture is constanly moving from the atmosphere to the surface of the earth and back again. At some stages in the cycle the moisture is more available for use than otners, but it is never destroyed. It is necessary at this point to make a distinction between water uae and water withdrawal. Water withdrawal is a rather straight!orward concept, and it is essentially that amount of water which is removed from a surface water or groundwater source. This is also known as water intake and says nothing about the use made of water. Water use is a more difficult term to define. Water use is not the same as water consumption yet they are related. To be consumed water must be used, yet not all water used is consumed. To be consumed water must be used in such a way that it is not readily available for subsequent users, al­ though it is physically undiminished. Consider, for example, a water user who withdraws 100 gallons of water from a water source. These 100 gallons may all be used in some industrial process but only 10 gallons may be consumed. The remaining 90 gallons are returned to a water source where they are again available for use. If a later user also withdraws 100 gallons and consumes 25, returning 75 gallons, the total water withdrawal will have been 200 gallons and the total use will have been 200 gallons. Consumption, however, will only have been 35 gallons. There is one further consideration which complicates the issue of water use and consumption. Water which is with­ drawn, minus that amount which is physically altered in form (for example, transfer to a vapor state) so as to make it less available, may be used and left in such a quality that it is virtually impossible to make further use of the water. Thus 6 water may be left unfit for some uses but not be actually physically altered in form (the original definition of con­ sumption) . Therefore, a further criteria for consumption may be necessary, one that takes into consideration changes in water quality that cause it to be unavailable for other uses, liowever, different types of activities are tolerant of dif­ ferent quality water. Thus water may be consumed, quality wise, for some types of uses but not for others. To compli­ cate matters even more, water used and left unfit for certain uses may, with passage downstream or by certain types of water treatment, again become fit for use (thus being renewed)• Types of Use There appear to be two major types of uses, those uses which internalize water in the production of a tangible good, and those uses which do not use water internally but use it as a median for the achievement of objectives. When water is used for industry, for agriculture, or for domestic purposes it is used for the production of some­ thing, although it is not necessarily consumed, either in form or in quality, in the process. Water in the factory may be used for the production of an automobile, on the farm for the irrigation of crops, and in the home for the growth of people. Other types of uses do not internalize water, nor do they alter it significantly as a result. The use of water for transportation, for recreation, or for aesthetic purposes 7 satisfies these criteria. Carried to extreme these types of uses may diminish the quality of the water for any of these purposes; excessive boating on a water body can decrease the quality of that particular resource for subsequent boaters. As a general rule, however, these uses of water do not dimin­ ish the water in quantity or quality, and thus impose no costs on subsequent users. For the purpose of this paper, the water "use" which will be considered will be water withdrawals for those uses which internalize water in its use. These types of uses will be restricted to domestic, municipal, and commercial uses, agricultural use for irrigation and for the production of livestock and related products, and water use in manufactur­ ing and power generation. The other uses of water, such as for transportation and for recreation, are recognized as equally important uses of water. The rationale for excluding such uses from this re­ search is two-fold. actly the same. First, the types of problems are not ex­ While all water use is related to people, those uses which use water "externally" are sufficiently dif­ ferent from those which use water "internally" to require special types of research. The second reason for excluding these uses from the research is that there are special agen­ cies whose main responsibility is directly concerned with transportation and recreation. While all uses of water must necessarily be related if a complete picture is to be 8 achieved, initial research is probably best done within the confines of a special research effort. Objectives Objective A The first objective is to identify spatial and tempo­ ral trends in the factors which are contributing most to the changing pattern of water use in Michigan. The factors which are responsible for broad, spatial differences in water use patterns are non-marginal. These have been identified in a number of different research ef­ forts, and are largely the characteristics of the population and the characteristics of economic activity. The focus of this objective will be on the historical development of the elements of population and economic activity which are most important in explaining changes in the amount and type of water use in Michigan over time. With both population and economic activity the most critical elements influencing water use will depend to a great extent on the scale of the study. If the area selected for study is extremely small the important elements will be different than if the study is oriented toward a larger area or toward a comparison between areas. Within any one region the amount of water used will be influenced by more detailed characteristics, but when the inquiry is the spatial pattern of use, as between regions, it is likely that detailed 9 information about individual regions will be less important than more general characteristics for which there are consid­ erable variations among regions. This is true of both population and economic activity. For example, the elements of population which may be impor­ tant when the object of study is a single city may be income, age structure, housing characteristics, etc. When the study is directed toward the difference between cities, however, the difference in the amount of water used may be determined so much by sheer differences in numbers of people that the detailed characteristics of the populations are either ob­ scured or become of much less importance. Similarly, the elements of economic activity important for determining the amount of water used in a single city may be a number of detailed characteristics, such as the general availability of water, or the pricing policy facing each pro­ ducer. When looking at the differences between cities or be­ tween regions factors such as these may become less important than the balance between agricultural or manufacturing acti­ vity, or the differences between types of agricultural or manufacturing activity. Objective B The second objective is to make tentative projections of selected water uses to 1980 and 2000 for Michigan and for regions thereof. The Michigan Water Resources Commission has recently 10 completed (19b8) an inventory of existing water use through­ out the state.^ Although the reports have several deficien­ cies, they are the most comprehensive treatment of water use on a state-wide basis that have ever been completed for Mich­ igan, and represent a relatively accurate picture of water use for 1967-1968. The contribution which this thesis can make to an un­ derstanding of water use in Michigan is an estimation of fu­ ture water use. There has been sufficient work done in this general area to determine that the use of water is going to increase. However, there have been no studies which have attempted to establish a spatial estimate of what this future use will be. Even with abundant resources of money and personnel it is very difficult to make an accurate projection of water use. And the difficulty increases in direct proportion to the de­ gree of detail which is attempted. rate projections have been made. For this study four sepa­ of these, three are directly related to the economic activities which are most important in influencing water use, manufacturing, power generation, and agriculture. The third projection which is presented repre­ sents a combination of the remaining types of water uses. For the purpose of this study it will be called domestic use, *Five separate regional reports (Southeastern Michigan, Lower Lake Huron, Lower Lake Michigan, Upper Peninsula, and Northern Lake Michigan and Lake Huron) from February, 1968 to December, 1968 (Lansingi Michigan Water Resources Commission)• 11 although it includes far more than just household uses. In­ cluded within this category, in addition to household uses, would be uses by commercial establishments, by institutions, by governments, and by other potential non-manufacturing users which might draw water from municipal water systems. The importance of these projections is not so much with the actual amount of water estimated to be withdrawn for use, but with the spatial patterns which develop. Therefore, as long as the estimates which are made are reasonable, and consistency is maintained between regions, the usefulness of these projections is not impaired. Objective C The third objective is to identify spatial patterns of future water uses which will be of importance to water plan­ ning efforts in the state. The focus of this objective is on a regionalization of the 1980 and 2000 water use estimates. This will be concerned not only with the total amount of water use, but with changes in the type of uses which can be expected among the several regions. It is possible for the state as a whole to exhibit a relatively continuous increase in water use, but for there to be quite a variation among the regions. Therefore, it is necessary that regional changes be identified if the state agencies concerned with water planning are to meet their re­ sponsibilities. 12 Scope In the consideration of a research effort the scope of the study is of great importance. With a given amount of re­ sources which can be expended for a study, the magnitude of the problem, either in geographic size or in complexity, will significantly influence the purpose of the research, the ob­ jectives, and the methods used. The geographical area covered by this research is state-wide. The examination of such a large area will neces­ sitate a more general approach than if the focus was on one county or even on a group of counties. Both the general the detailed approaches have merit. The detailed study of a small area will enable the identifi­ cation of important variables, and the magnitude of their in­ fluence, to be more accurate, and as a result any quantita­ tive estimates of water use at a future date will undoubt­ edly be more precise. The accuracy of such a study, however, does not indicate the nature of the situation for the larger area, a region or a state. There are important spatial con­ siderations that can be understood only when the complete situation can be realized. In fact, a thorough understand­ ing of water use relationships of the small area can often be seen only within the context of the larger area. The manner in which data is presented, the size of the enumeration tract, is important with regards to the unity of the data. By reporting data in small tracts it is possible 13 to reorganize it to conform to any number of regional areas. When the data is presented on a large area basis the flexi­ bility of the data becomes less. And, data for small areas is often hidden within the larger ones. For this research the preliminary data and resulting water use estimates have been organized according to a re­ gionalization of counties which was recommended by the Mich­ igan Water Resources Commission (see the map on page 14). Regions I, II, III, IV, and V represent the groupings of coun­ ties which most closely correspond to major watersheds of the state. Regions II and III have each been divided into parts A and B because of the relatively large size of regions II and III. In estimating future conditions of water use the time framework is always a significant consideration. Some bal­ ance must be achieved between the extreme long range pro­ jections which are likely to be greatly in error because of the long span of time, and very short range projections which, although more precise, are less useful because of the re­ stricted time period. For this study 1980 and 2000 were cho­ sen as target dates. These were selected completely arbi­ trarily, and represent a compromise between short and long range projections. Introduction to Methodology The general research plan for the study is to iden­ tify the principle factors responsible for water use, to ranij tawaw m IftM LVCf iat-aaa KMMUWr ■tf ioicoa* t m x m m M M OL* CLM I H (M T « I M I I l U l m i n ,! U aa*Ti»T 1 1***11** IONM (HIITfai1 cat •«***■* — - — - " — T c«at mu Figure 1. Planning Regions li> project tlie magnitude of these factors to 1980 and 2000, and from these to estimate selected future water uses. A more detailed description of the methods used in making these estimates will be found in later chapters. However, there is one point which much be discussed at this tine since it pro­ vides a basis for the rest of the study. This most important topic involves the nature of forecasts and projections. Forecasts and Projections In its most elementary form a forecast is nothing more than a statement about the condition in which something is expected to be at some future time. Such forecasts can be made on a continuum which ranges from pure guesses to those which are based on sophisticated models which rely upon per­ fect knowledge of future conditions of those factors which are responsible for the thing which is being examined. At one end of the continuum, the guess, forecasts are generally rejected in favor of a method which has more pre­ dictive power. At the other extreme, the detailed knowledge which is needed if the relevant factors associated with the object under investigation is seldom present. Consequently, forecasts are generally based upon a compromise between the "guess'* and perfect knowledge. Most "compromise" forecasts rely to a great extent upon the observance of the past history of the object of the forecast and of the factors which are responsible for it. These compromise forecasts which utilize data on past lb conditions arc commonly called projections. Kuzncts defines projections as "...statements about the future that claim to be derived from empirically tested propositions concerning the past."* The Mature of Projections 2 Statements about the future which are based upon the past appear at first to be simple logic; since the future can never be known by the observation of empirical evidence it would seem realistic to assume that the events of the future would resemble those of the immediate past. Yet there are two criteria which must be met before such a statement can be valid. First, the past and the future must be interrelated in some logical manner. And second, the past events must have some semblance of order about it. The first criteria, that of relationship between the past and the future, is an obvious necessity. Unless this relationship is present any attempt to assess the future would be in vain. Zt would be futile to expect that the a- mount of water used in Michigan in 1980 would be in any way related to the trends in the ratio of male-female births. However, it would be quite realistic to expect that changes *Simon huznets, "Concepts and Assumptions in Long-Term Projections of National Product," Studies in Income and Wealth, Vol. 16 (Washington; National bureau of economic Researcli, 1954) , p. 9. 2 This section is based upon the points developed in Kuznets, Ibid. 17 in population levels and economic activity might have an effect. The second criteria which must be met before a projec­ tion can be made is that there must be some systematic order­ ing of past events. If the relevant factors have occurred randomly in the past there is no reason to expect that they will occur in any other way in the future. And it would be impossible to project a random past in any manner except ran­ domness, a useless effort if the purpose is to determine what the future may hold. Therefore, unless past conditions are arranged in some meaningful pattern it will be useless to make projections based upon them. exists in past data can be tested. The extent to which order Therefore, it is possible to proceed with the realization of the probable validity of the projection. The question of prediction versus forecast is an im­ portant one. The principle distinction is related to the existence of a causal relationship between the object of study and one or more associated variables. A prediction im­ plies that one or more factors cause another factor to assume a certain condition. It assumes that specific magnitudes of A, B, and C will cause 0 to exist in a certain form. A forecast does not necessarily involve the function of causality, yet may equally well approximate the future. physicist and an illiterate may both throw a stone into the air and state that it will fall to earth. The physicist, to A 18 the extent that he applies the concept of gravity, ing that the stone will fall. is predict­ The illiterate may have no un­ derstanding of physics, but from past experience believes that the stone will behave as previous stones have behaved. then is making a forecast and not a prediction. He In a socio­ logical example, there is a strong correlation in Michigan between a county's total population and its population den­ sity. Thus one may make a statement about future population levels based upon trends in population density. This does not, however, involve a complete explanation since the matter of the area of the county has not been considered. The extent to which either prediction or estimation is satisfactory will depend to a large extent on the purpose of the projection. If the object is to develop an understanding of why and how the causal variables operate one must work toward prediction. On the otherhand, if the intent is merely to determine the magnitude of a certain condition, an expla­ nation may not be necessary. If it was possible each year to accurately forecast the first snowfall by the date of the last snowfall in the previous year this would be satisfactory if the object was just to determine the date of the first snowfall and not provide an explanation of why. Such a spurious relationship is seldom, if ever, found. Most estimations as opposed to predictions are based upon re­ lations which are obvious but are not well enough understood to be fully explained, or for which the lack of data is a iy restricting factor. usefulness. This, however, does not prevent their If the object is to estimate how much timber will be cut in the nation next year, and this can be done by projecting the quantity cut in previous years, it is not nec­ essary to know what the demand for housing, furniture, etc. will be. Often the addition of successive factors will increase the ability to forecast the future with accuracy. To the ex­ tent that data is available that permits this, this should be the goal. When data are lacking forecasts must be made with data at hand. The present research must, of necessity, be more of a forecast than of a prediction. The detailed, historical data which is required for the more complex projection methods can not be obtained and, therefore, these methods must be passed over in favor of methods which can utilize the data which is available. The forecasts of water use which are presented are based, however, on sound relationships which have been tested in other research which has had access to the type and quality of input data which is necessary to utilize the more sophisticated methods. Assumptions Forecasts of the future must necessarily involve as­ sumptions about the future. There are certain basic condi­ tions which must be stabalized if any projections can be made. It must be assumed that water availability will re­ main the same, and will play essentially the same role that it has played in the immediate past. There have been many schemes to increase the water in certain areas of the United States. The North American Water and Power Alliance and weather modification are examples. In Michigan, there has been discussion of a bake Michigan to Lake Huron canal which would provide increased water for some areas of the state and undoubtedly influence the spatial pattern of water use. In addition, there are a few potential reservoir sites which could possibly be utilized to increase the available water. It is assumed that these developments will not take place to any appreciable extent prior to 2000. Further, it is assumed that there will be no major technological "breakthrough" which substantially lowers the per capita demand for water. Equally restrictive, it must be assumed that there will be no major droughts in 1980 or 2000. The "normal” available water must be considered as existing in the target years. A second assumption is that there will be no major war or large inflationary or deflationary periods in the economy in the years 1980 and 2000. Such wars or economic changes may very well occur prior to these target years, but it must be assumed that by 1980 and 2000 their effect will have been ameliorated so that it is no longer significant. There are several other assumptions which must be made with regards to each of the domestic, industrial, and 21 agricultural projections. spective chapters. These will be discussed in the re­ CHAPTER II FACTORS CONTRIBUTING TO DOMESTIC, MUNICIPAL, AND COMMERCIAL WATER WITHDRAWALS Introduction The category of water use labeled "domestic, municipal, and commercial" is a very important one in that it is one of the necessary elements of a modern urban system. This cate­ gory of water use includes three principle types of with­ drawals: dences, (1) domestic - water withdrawals by private resi­ (2) municipal - withdrawals for use in schools, hos­ pitals, and other urban services, (3) commercial - laundries, restaurants, car washes, etc. The purpose of this chapter will be to examine the factors behind differences in domestic, municipal, and com­ mercial withdrawal rates among different parts of an area. No attempt will be made to relate these to the Michigan sit­ uation in particular because of the inability to handle such detail on a state-wide scale, but to just discuss in general the findings of related research. Composition of Withdrawals Per capita domestic, municipal, and commercial water 22 23 withdrawal rates combined have increased considerably in the past and are expected to increase further in future years. The Mater Resources Council has estimated that, nationally, the daily per capita withdrawal from municipal systems for all non-industrial uses will rise from 121 gallons in 1965 to 123 gallons in 1980 and 125 gallons in the year 2000.^ Of the total withdrawals, the amount required for com­ mercial uses is expected to remain constant at 28 gallons per capita per day. The amount required for public or municipal purposes is expected to decrease from 20 gallons per capita per day in 1965 to 18 in 1980 and 16 in the year 2000. Thus the total category increase in per capita withdrawals is ex­ pected to result from increased domestic use. For the total Great bakes region this is expected to increase from 73 gal­ lons per capita per day in 1965 to 77 gallons in 1980 and 81 in the year 2000.2 Elements of domestic Withdrawal Increases The majority of the research which has been done on residential use points to three factors as being primarily responsible for the increased per capita use ratest (1) the development of modern home appliances which utilize large amounts of water, (2) increased incomes which allow the ^water Resources Council, The Nation's Water Resources (Washington: U.S. Government Printing Office, 1968), p. 4-1-2. 2 Ibid. 24 purchase of the aforementioned appliances, (3) the movement of larger numbers of people to the suburbs. Modern Appliances There have four appliances which have been respon­ sible for increasing the per capita withdrawal rates for d o ­ mestic purposes. These are clothes washers, dishwashers, ga r ­ bage disposals, and air conditioners. Depending on the type of washing machine the mean rate of water use is between two and seven gallons per minute. The normal load requires from 36 to 50 gallons.^ The higher rates are found in the automatic washers, and the trend toward the use of automatic washers suggests that increasing amounts of water will be required for this purpose. dishwashers. The same general pattern is found in automatic An Ohio study found that the amount of water used for a single washing in an automatic washer was approxi­ mately 40 per cent greater than when the washing was done by hand. 2 This survey was taken in 1956 and it was suggested that modern washers probably use more water than did the old­ er models. A study made in St. Louis indicated that homes with air conditioners used between four and one-half and sev­ en times as much water in the summer as in the winter. In *J.G. C a m s , "Service Lines and Meter Requirements of Domestic Water Connected Devices," Journal of the American Water Works Association, Vol. 58, NoT T5 (October, 1966), pTT25T6-----------------2 Rudolfo Silva, "Land Uses and Water Consumption Re ­ quirements," Public Works, Vol. 90, Ko. 4 (April, 1959), p. 123. 25 homes without air conditioning summer use was less than one and one-half times the winter rate.^ Increased Incomes Increased incomes allow the populus to enjoy the fruits of technology, generally resulting in increased per capita withdrawal rates. In a study of per capita withdrawal rates in Illinois, a strong relationship was found between family income and per capita use rates. Families with high incomes used, on the average, five times as much water per person as did families with low incomes. In addition to home appliances, higher family incomes have allowed the construc­ tion of homes with more complete sanitary facilities, includ­ ing multiple bathrooms. Even home swimming pools are be­ coming commonplace. A more detailed study of this Rhode Island. type of use was made in The purpose of the study was to see if the amount of water used in households was related to social sta­ tus. The significance of a number of variables proporting to represent social status were tested. Only three of these, however, were found to be significantly correlated with the amount of household water use. These variables were ^Kenneth S. Watson, "Water Requirements of Dishwashers and Food Waste Disposers," Journal of the American Water Works Association, Vol. 55, No. 5 Ulay, IH-J), pp. 5^5-^59. 2 Brent O. Larson and li.E. Hudson, "Residential Water Use and Family Income," Journal of the American Water Works Association, Vol. 43, No. 7 (July, 1951), p. 6lo. 2t> household income, house value (which is directly related to income), and household size.^ Other social variables have also been found to be re­ lated to residential water use. In a study in Kankakee, Il­ linois, assessed valuation of the house, family income, edu­ cation, and occupation were statistically correlated with water use per household. 2 It is interesting to note that all of these variables are interrelated to a certain extent. Suburban Living The trend toward suburban living has accelerated the per capita water withdrawal rates for residential use. This is overwhelmingly the result of larger lawns and the demand for water to irrigate them. In fact, the increase in the de­ mand for lawn sprinkling has changed the design of water sys­ tems in surburban areas. With the pre-World War II condi­ tions most water systems were designed with the controlling factor being the quantity of water needed in case of a bad fire. In many suburban areas it has become necessary to make the potential lawn sprinkling requirements the controlling factor. Demands for any fire load are almost always covered ^Irving A. Spaulding, Household Water Use and Social Status (Kingstons Agricultural Experiment Station, University of Rhode Island, 1967), pp. 7-25. 2 Dorthy F. Dunn and Thurston E. Larson, wRelationships of Domestic Water Use to Assessed Valuation, with Selected Demographic and Socio-Economic Variables," Journal of the American Water Works Association, Vol. 55, No. 3 (April, 1963), pp. 441-449. 27 under the sprinkling capacity.^ The tremendous growth rate of the suburbs almost in­ sures that this trend will continue. Between 1950 and 1955, 98 per cent of the population increase was in metropolitan areas, and the suburban growth rate was seven times as great 2 as that of the central cities. Although these are national data this trend is also evidenced in Michigan (see Table 5)• The amount of water required for watering lawns will not be uniform throughout a large area because of differences in climate, largely temperature and precipitation. Even in Michigan there is sufficient variety in climatic patterns to cause a potential difference in sprinkling requirements in different parts of the state. Figures 2 and 3 indicate the situation for the bower Peninsula. The map of potential evapotranspiration indicates how much moisture would be evap­ orated and transpired if there was sufficient moisture avail­ able. The values for potential evapotranspiration generally diminish toward the northern part of the state. The map of normal moisture deficit shows the difference between the maximum possible evapotranspiration and the actual evapotranspiration (given adequate moisture) (which may be less than the maximum possible because of inadequate soil moisture). The greater the deficit the more critical is the lack of moisture. ^Angus D. Henderson, " The Lawn Sprinkling L o ad,” Journal of the American Water Works Association, Vol. 48, Mo. J (' A pril, 1955), pp. 361-362.---------------2 p. 121. Silva, "Land Uses and Water Consumption Requirements," Figure 2. Normal Potential Evapotranspiration a Id aA. Stephen Messenger, "The Water Balance of the Lower ninsula of Michigan," Papers of the Michigan Academy of ience, Arts, and Letters, V o l . XLVII (1962), p~! 586. ^Millimeters AII4U M N III 9»CI0lA I ct*it 1 A ItI MAC OCtANA .««»AT*0 'm c COITA i U»«f.‘ HINT r iOAIA .CLIATOH 1 CALHOUN r ilMiti r Figure 3. Normal Deficit A. Stephen Messenger, "The Water Balance of the Lower Peninsula of Michigan," Papers of the Michigan Academy of Science, Arts, and Letters, Vol. XLVII (1962), p. 387. ^Mi 11 ime 16r s 30 It can be seen that there are definite spatial differences, at least in the bower Peninsula, in the soil moisture deficit. Pricing Policies Although the withdrawal rates for domestic, municipal, and commercial uses have risen in the past and are expected to rise considerably in the future, the rate of increase is probably lower than it would have been without the operation of a very basic economic principal. This principal is the price elasticity of demand for water and the tendency for less water to be used when the cost of that water increases in certain kinds of situations. a number of studies have shown empirically that this process actually works in the area of water utilization. In a study of municipal water systems throughout the nation a direct relationship was found between water use in urban areas and the price of water. 2 When SO per cent of the water was metered there was no firm trend in the reduction of use rates. In the 50 to 95 per cent metered category there was a sharp decrease in the per capita use rates six per cent reduction in per capita u s e ) . (approximately a This trend was ^For a most complete discussion of the water balance concept see C.W. Thornthwaite and J.R. Mather, The Water Balance (Centerton. New Jersey: brexel Institute ot technology, l 9 in>) . 2 Ralph Porges, "Factors Influencing Per Capita Water Consumption," Water and Sewage Works, Vol. 104, No. 5 (May, 1957), pp. 199-264. 31 accentuated in the over 99 per cent metered category. On the basis o£ these data the savings in water would be about 25 per cent when going from an unmetered system to 100 per cent metering. In an analysis of the water use statistics of 441 cities as published by the American Water Works Association a close correlation was found between the price of water and the per capita use rate. A summary of their findings is pre­ sented in Table 2. Table 2 Price Elasticity of Residential Water Deroanda Cost of Water (c/1000 ga l . ) Per Capita Use (gal. per day) .70 .60 .50 .40 .30 .20 .10 137 143 154 178 205 246 296 aH. Seidel and E. Baumann, *A Survey of Operating Data on Water Works in 1955," Journal of the American Water Works Association, Vol. 59, No. 5 (May, 1967), p. SS!j . Figure 4 shows the demand-price relationship for water use for domestic purposes in western cities. This data again substantiates this economic relationship. Despite these apparently strong relationships overall, it is often difficult to identify them when dealing with any 32 70 50 o 40 > ' 30 *>\