1|Page

Course: Water Supply and sanitation engineering

Chapter 2: Population Forecasting

Compiled by:

M.M.Qawdhan

(MSc Water Resources Engineering, 2019)

Compiled by M.M.Qawdhan 2019

2|Page

2. POPULATION FORECASTING
Design of water supply and sanitation scheme is based on the projected population of a
particular city, estimated for the design period. Any underestimated value will make system
inadequate for the purpose intended; similarly overestimated value will make it costly.
Changes in the population of the city over the years occur, and the system should be designed
taking into account of the population at the end of the design period.

Factors affecting changes in population are:

 increase due to births
 decrease due to deaths
 increase/ decrease due to migration
 increase due to annexation.

The present and past population record for the city can be obtained from the census
population records. After collecting these population figures, the population at the end of
design period is predicted using various methods as suitable for that city considering the
growth pattern followed by the city.

2.1. ARITHMETICAL INCREASE METHOD

This method is suitable for large and old city with considerable development. If it is used for
small, average or comparatively new cities, it will give lower population estimate than actual
value. In this method the average increase in population per decade is calculated from the
past census reports. This increase is added to the present population to find out the population
of the next decade. Thus, it is assumed that the population is increasing at constant rate.
Hence, dP/dt = C i.e., rate of change of population with respect to time is constant.

Therefore, Population after nth decade will be Pn= P + n.C (1)

Where, Pn is the population after ‘n’ decades and ‘P’ is present population.

Compiled by M.M.Qawdhan 2019

3|Page

Example: 1

Predict the population for the year 2021, 2031, and 2041 from the following population data.
Year 1961 1971 1981 1991 2001 2011
Population 8,58,545 10,15,672 12,01,553 16,91,538 20,77,820 25,85,862

Solution

Year Population Increment

1961 858545 -
1971 1015672 157127
1981 1201553 185881
1991 1691538 489985
2001 2077820 386282
2011 2585862 508042
Average increment = 345463

Population forecast for year 2021 is, P2021 = 2585862 + 345463 x 1 = 2931325
Similarly, P2031 = 2585862 + 345463 x 2 = 3276788
P2041 = 2585862 + 345463 x 3 = 3622251

2.2. GEOMETRICAL INCREASE METHOD (OR GEOMETRICAL PROGRESSION

METHOD)

In this method the percentage increase in population from decade to decade is assumed to
remain constant. Geometric mean increase is used to find out the future increment in
population. Since this method gives higher values and hence should be applied for a new
industrial town at the beginning of development for only few decades. The population at the
end of nth decade ‘Pn’ can be estimated as:
Pn = P (1+ IG/100) n (2)
Where, IG = geometric mean (%)
P = Present population
N = no. of decades.

Example: 2

Compiled by M.M.Qawdhan 2019

4|Page

Considering data given in example 1 predict the population for the year 2021, 2031, and 2041
using geometrical progression method.

Solution
Year Population Increment Geometrical increase
Rate of growth
1961 858545 -
1971 1015672 157127 (157127/858545)
= 0.18
1981 1201553 185881 (185881/1015672)
= 0.18
1991 1691538 489985 (489985/1201553)
= 0.40
2001 2077820 386282 (386282/1691538)
= 0.23
2011 2585862 508042 (508042/2077820)
= 0.24

Geometric mean IG = (0.18 x 0.18 x 0.40 x 0.23 x 0.24)1/5

= 0.235 i.e., 23.5%
Population in year 2021 is, P2021 = 2585862 x (1+ 0.235)1 = 3193540
Similarly for year 2031 and 2041 can be calculated by,
P2031 = 2585862 x (1+ 0.235)2 = 3944021
P2041 = 2585862 x (1+ 0.235)3 = 4870866

2.3. INCREMENTAL INCREASE METHOD

This method is modification of arithmetical increase method and it is suitable for an average
size town under normal condition where the growth rate is found to be in increasing order.
While adopting this method the increase in increment is considered for calculating future
population. The incremental increase is determined for each decade from the past population
and the average value is added to the present population along with the average rate of
increase.
Hence, population after nth decade is Pn = P+ n.X + {n (n+1)/2}.Y (3)
Where, Pn = Population after nth decade
X = Average increase
Y = Incremental increase

Compiled by M.M.Qawdhan 2019

5|Page

Example: 3

Considering data given in example 1 predict the population for the year 2021, 2031, and 2041
using incremental increase method.

Solution
Year Population Increase (X) Incremental increase (Y)
1961 858545 - -
1971 1015672 157127 -
1981 1201553 185881 +28754
1991 1691538 489985 +304104
2001 2077820 386282 -103703
2011 2585862 508042 +121760
Total 1727317 350915
Average 345463 87729

Population in year 2021 is, P2021 = 2585862 + (345463 x 1) + {(1 (1+1))/2} x 87729
= 3019054
For year 2031 P2031 = 2585862 + (345463 x 2) + {(2 (2+1)/2)} x 87729
= 3539975
P2041 = 2585862 + (345463 x 3) + {(3 (3+1)/2)} x 87729
= 4148625
2.4. GRAPHICAL METHOD
In this method, the populations of last few decades are correctly plotted to a suitable scale on
graph (Figure 5.1). The population curve is smoothly extended for getting future population.
This extension should be done carefully and it requires proper experience and judgment. The
best way of applying this method is to extend the curve by comparing with population curve
of some other similar cities having the similar growth condition.

Compiled by M.M.Qawdhan 2019

6|Page

Figure 5.1 Graphical method of population forecasting

2.5. COMPARATIVE GRAPHICAL METHOD

In this method the census populations of cities already developed under similar conditions are
plotted. The curve of past population of the city under consideration is plotted on the same
graph. The curve is extended carefully by comparing with the population curve of some
similar cities having the similar condition of growth. The advantage of this method is that the
future population can be predicted from the present population even in the absence of some
of the past census report. The use of this method is explained by a suitable example given
below.

Example: 4

The populations of a new city X given for decades 1970, 1980, 1990 and 2000 were 32,000;
38,000; 43,000 and 50,000, respectively. The cities A, B, C and D were developed in similar
conditions as that of city X. It is required to estimate the population of the city X in the years
2010 and 2020. The population of cities A, B, C and D of different decades were given below:

(i) City A: 50,000; 62,000; 72,000 and 87,000 in 1960, 1972, 1980 and 1990,
respectively.
(ii) City B: 50,000; 58,000; 69,000 and 76,000 in 1962, 1970, 1981 and 1988,
respectively.
(iii) City C: 50,000; 56,500; 64,000 and 70,000 in 1964, 1970, 1980 and 1988,
respectively.

Compiled by M.M.Qawdhan 2019

7|Page

(iv) City D: 50,000; 54,000; 58,000 and 62,000 in 1961, 1973, 1982 and 1989,
respectively.

Population curves for the cities A, B, C, D and X are plotted (Figure 5.2). Then an average
mean curve is also plotted by dotted line as shown in the figure. The population curve X is
extended beyond 50,000 matching with the dotted mean curve. From the curve, the
populations obtained for city X are 58,000 and 68,000 in year 2010 and 2020.

100
Population in thousand

80

60

40

20

0
1960 1970 1980 1990 2000 2010 2020

Year
Population curve

Figure 5.2 Comparative graph method

2.6. MASTER PLAN METHOD

The big and metropolitan cities are generally not developed in haphazard manner, but are
planned and regulated by local bodies according to master plan. The master plan is prepared
for next 25 to 30 years for the city. According to the master plan the city is divided into
various zones such as residence, commerce and industry. The population densities are fixed
for various zones in the master plan. From this population density total water demand and
wastewater generation for that zone can be worked out. By this method it is very easy to
access precisely the design population.

2.7. LOGISTIC CURVE METHOD

This method is used when the growth rate of population due to births, deaths and migrations
takes place under normal situation and it is not subjected to any extraordinary changes like
epidemic, war, earth quake or any natural disaster, etc., and the population follows the growth

Compiled by M.M.Qawdhan 2019

8|Page

curve characteristics of living things within limited space and economic opportunity. If the
population of a city is plotted with respect to time, the curve so obtained under normal
condition looks like S-shaped curve and is known as logistic curve (Figure 5.3).

Saturation Population, Ps

M
Point of L
inflexion

K
Curve of
J growth rate

Figure 5.3 Logistic curve for population growth

In Figure 5.3, the curve shows an early growth JK at an increasing rate i.e. geometric growth
dP dP
or log growth, ∝ P, the transitional middle curve KM follows arithmetic increase i.e. =
dt dt
constant. For later growth MN the rate of change of population is proportional to difference
dP
between saturation population and existing population, i.e. ∝ (P s - P). A mathematical
dt

solution for this logistic curve JN, which can be represented by an autocatalytic first order
equation, is given by

loge ( ) - log (
Pc–P
e
Pc–P0
) = -K.P .t
s (4)
P P0

where, P = Population at any time t from the origin J

Ps= Saturation population

P0 = Population of the city at the start point J
K = Constant
t = Time in years

Compiled by M.M.Qawdhan 2019

9|Page

From the above equation we get

loge = - K.Ps.t (5)

After solving we get,

Pc
P= 1+
Pc —P0
Sog—1(–K.Pc.t) (6)

Po e

Pc–P0
Substituting = m (a constant) (7)
P0

and - K.Ps = n (another constant) (8)

Pc
we get, P= (9)
1+ nlog—1
e (n.t)

This is the required equation of the logistic curve, which will be used for predicting population.
If only three pairs of characteristic values P0, P1, P2 at times t = t0 = 0, t1 and t2 = 2t1 extending
over the past record are chosen, the saturation population Ps and constant m and n can be
estimated by the following equation, as follows:

2P0P1P2 –P2(P0 + P2)

Ps = 1
2 (10)
P0P2 –P1

Pc–P0
m=
P0

2.3 P0(Pc–P1)
n= log (11)
10
t1 P1(Pc –P0)

Example: 5
The population of a city in three consecutive years i.e. 1991, 2001 and 2011 is 80,000;
250,000 and 480,000, respectively. Determine (a) The saturation population, (b) The equation
of logistic curve, (c) The expected population in 2021.

Solution

It is given that

Compiled by M.M.Qawdhan 2019

10 | P a g e

P0 = 80,000 t0 = 0

Compiled by M.M.Qawdhan 2019

11 | P a g e

P1 = 250,000 t1 = 10 years

P2 = 480,000 t2 = 20 years

The saturation population can be calculated by using equation

2P0P1P2 –P2(P0 + P2)

Ps = 1
2
P0P2 –P1

2 x 80,000 x 2,50,000 x 4,80,000–2,50,000 x 2,50,000 x (80,000 + 4,80,000)

=
80,000 x 4,80,000 –2,50,000 x 2,50,000

= 655,602

Pc–P0 655,602–80,000
We have, m = = = 7.195
P0 80,000

2.3
n= log P0(Pc–P1)

10
t1 P1(Pc –P0)

2.3
= log
10 10

= -0.1488

Population in 2021

Pc
P=
1+ nlog—1
e (n.t)

6,55,602
=
1+ 7.195 x log—1
e
(—0.1488 x 30)

6,55,602
= = 605,436
1+ 7.195 x 0.0117

Compiled by M.M.Qawdhan 2019

12 | P a g e

Questions

2.7.1. Explain different methods of population forecasting.

2.7.2. The population data for a town is given below. Find out the population in the year
2021, 2031 and 2041 by (a) arithmetical (b) geometric (c) incremental increase
methods.

Year 1971 1981 1991 2001 2011

Population 84,000 1, 15,000 1, 60,000 2, 05,000 2, 50,000

2.7.3. In three consecutive decades the population of a town is 40,000; 100,000 and
130,000. Determine: (a) Saturation population; (b) Equation for logistic curve; (c)
Expected population in next decade.

Compiled by M.M.Qawdhan 2019