334 THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY. CHINA
SD MODEL OF URBAN WATER USING OF TIANJIN
Li Puhua and Tang Chunpeng
Tianjin Research Institute of Environmental Protection
ABSTRACT
The paper presents a preliminary attempt for application of system dynamics
in the study of urban water using and its management.
The SD model of urban water using describes the complicated system behavior.
It will help us to understand and trace time interval behgavior of urban
water using system, to observe and handle dynamic trend of system behavior,
to obtain the features of urban water using in different time. Therfrom we
may have a quantified rational knowledge about water demand and have to find
an effective may to strengthen scientific management of urban water using, to
promote rationalization of urban water using,to save water resources and to
resolve unbalance between supply and demand. Meanwhile, necessary informations
and quantized references should be provided for administration and depart—
ments concerned to determine the speed of economic development, study layout
and structure of industrial development, adjust key points and methods for
urban water using.
1. Raising of The Problems
At present, like the situation in many other countries in the world, crisis
of fresh water resources in China is growing seriously with each passing day,
especially the shortage of urban water resources is a pressing problem. Water
supply has become an important restrictive factor in the development of na~
tional economy. How to resolve the unbalance between supply and demand of
urban water resources ar how to bring the limited water resources into full
play for maximum economic benefit? These are spectacular important problems,
which need to be solved urgently. On the basis of a series of scientific re-
scarches, we try to reflect these problems from one side and lock for an
effective way to resolve the unbalance between supply and demand of urban
THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY. CHINA 335
water resources under the premise of limited water resources.
As a result of comprehensive review on water using in the whole city of
Tianjin, we considered it is necessary to establish a related model for urban
water using in Tianjin for further analyzing urban water using in every
aspect and different time period, simulating urban water using in past and
at present, tracing urban water using in futures in order to more deeply,
systematically and accurately master the objective law of urban water using
and thoroughly analyze the propect of urban water using in future. A rather
clear picture of urban water using in Tianjin in the past, present and fu-
ture thus can be drawn out. Corresponding measures will he teken according to
this picture, every effort will be made to regulate, control and manage the
mode of urban water using and to improve water using and management level to
a possible extent under available condition of water resources. Development
of national economy with high speed to the target———-2000 AcD. can be gua~
ranteed, as well as rational using of water, scientific management, saving
of fresh water resources and making the most of economic benefit of water.
For the above-mentioned we decided to establish a related model for urban
water using in Tianjin.
il. Features of Urban Water Using System and Applicability of SD Method
The system of urban water using is a complicated social system, goal of which
is pluralistic, variable and of remarkable non-solitary sharacter, For exam—
ple, index of each quantity of urban water using is not a single one, but a
set of index system of combined measurement. Generally, parameters and struc—
ture in-the system of urban water using vary with time, the change of parame-
ters in different years van be observed directly, while the change of struc~
ture can be reflected only from the difference of interaction relationship in
various stages. Moreover, this social system in non-linear, so it often shows
the non-additivity of action and rusult. Even measures were adopted for water
saving in double, but the result of water saving may not increase in double.
This feature showed that we can hardly describe the system with mathematic
mode» In short, features of urban water using system can be described as
follows: 1. system is large in scale, it is not limited in some simple ‘
336 THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY, CHINA
element, but a macroscopic phenomemon of various water quantity and condi-
tions of water using in the whole urban water usings 2. complicated in struc—
ture, in which not only exists relation between water quatities, but also
relation between action of man's administration and water, as well as more
complicated relation between various water quantities and new growing infor~
mations;3. multi-functions, so as mentioned above, the goal of system is
pluralistic and non-solitary; 4. mlti-factors, in the system there are not
only various factors of water, but also man's factor (such as management
factor), nature factor and eto.
As regards to such complicated system, decision study of urban water using
needs to be quantified and dynamiciztd, it may not only rely on ordinary
means and methods traditional inferential calculation and qualitative analy-
sis, but on the development of static model turhing toward to dynamic model.
Besides, urban water using system is a system, deficient in data by its na-
ture, moreover, there are many affecting factors, therefore, it is unsuf-
ficient, uncomplete and unactual to describe the system only according to
data. Many features, going to be described in the system, can not-be directly
expressed numerically, they are of-a definite fuzzy feature,for example, the
delay phonomemon in the system.
In accordance to these features corresponding requirements were proposed for
the method with model, i es the'model building might be installed deeply
into the interior of the system in order to analyze the structure of system,
reveal interdependent relation between various elements and interact on each
other in the system, the content expressed by model should be kept in a same
strudture relation with the real system to a certain degree, close to or
coincident with the actual circumstance in past and at pressent, possibility
of development in future. The whole dynamic process flowing in past, at
present and in future should be understood from the view.of macroscopic ten-
dency, and the feature of urban water using in each time will be shown and
studied, i, es it is possible to describe time interval behaviour of the
system in a state space. Because the history is continuout, one can follow
the law of its development in past, present and future; on the basis of the
. past and present, meanwhile with utilization of a logic structure, we can
link them up with the future, analyze those uncertain factors by means of
systematic, moving, coneective and changeable philosophic view to describe
and reveal the actual system being studied. In the meantime, because human
THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY. CHINA 337
ability falls behind his will in dynamic tracing of complicated social
system during long time period, so it is necessaty to build an indispensable
man-machine system, wkibh may accomplish the tracing to the social system.
According to the above-mentioned, a macroscopic, fuzzy and long-term trend
study of. non-linear, complicated system with time delay can not be solved
only by mathematical analysis. An alternative is the experimental method for
system analysis, i. e. a system simlation method. System Dynamics is just
a new scientific subject in this field.
Our extra attention was directed to the specific predominance of SD method,
it permits mode builder to form various modes of system performance through
the modification of parameters and repetitive simulation after different
policy choice,providing the convenience for the study of policy effect.
Through mathematical simulation for the system, tracing the conceived policy
scheme and observing the changing trend of Bhe system, it may not only make
the actual policy decision more coincident with the objective law of the
system development, but also affect the thinking activity of policy decider
in return.
The more features of the System Dynamics method will be revealed and ana—
lyzed in detail and profound, the more distinctively will be shown the
method availability for urban water using system. Thus, it impeled us to
build a SD model for urban water using in Tianjin, a attempt in appli~
cation. of System Dynamics is expected.
III. SD Mosel of Urban Water Using in Tianjin
Process of model building is a process by use of a special language to
describe the models of system behavior and by use of computer to solve
this dynamic mode progressively, that is jte say, a real system through a
specially designated abstraction can be transferred on computer to a arti-
ficial system, which can be controlled and regulated. The steps arc as
follows:
1. Determine goal of the system and aim of model building
The SD model of urban water using oan desoribe the complicated system be—
havior ‘in a long time interval of 40 years from 1960 to 2000. With the aid
338 THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY. CHINA
of computer model simulated qualitative and quantitative conditions of urban
water using system within these 40 years, so as to complete abstractive pro-
cess of the real system. A series of simlation results and different data
@f variables obtained after computer operation will help us to understand
and trace time interval behavior of urban water using system reflected by
model (including the course of history, which has passed or will pass in
future), to observe and handle dynamic trend and objective law of the sys-
tem behavior of urban water using, to obtain the features of urban water
using in different time, especially in 2000 A.D. Therefrom we may have a
quantized rational knowledge about water demand produced due to the deve—
opment of national economy and improvement of people's living standard. In
the meantime, we shall have a vivid understanding and more imaginal, quan
titative concept about hardship and sense of crisis due to the fact that
the limited water resources are attempting to meet such demand» We have to
find an effective way to. strengthen scientific management of urban water
using, to promote rationalization of urban water using, to save water re—
sources and to resolve unbalance between supply and demand. Meanwhile, nece-
ssary informations and quantized refernces should be provided for admini-
stration and other departments concerned to determine the speed of economic
development, study layout and structure of industrial development, adjust
key points and methods for urban water using, strengthen scieutific mana—
gement of industry water, reform technology of water using, make an invest-
ment in fixed assets for water saving and etc.
2. Determine the boundary and range of the system
Because the built model is used to describe urban water using system, so
all quantities of water (such as gross water, water taking, saved water,
recycling utilization water and etc.), fixed assets (such as for water
saving, technology reform and etc.), management level, pressure of water
saving, reoycling rate and etc. are regarded as internal variables entering
into system boundary, while values of industry output, quantity of water
resources and etc.—--exogenous variables.
3. Determine and select model variables and causal relations between vari—
ables
Selection of model variables must be carried out around the goal of system
and the main features of the system might be shown through selected vari-
THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY. CHINA 339
ables. Build of feedback relation link is the key of model building, it
desoribes internal structure of the real system, it is a frame of system
structure. The causal relations of this model is shown in Fig. 1. The main
feedback loops in this model are shown in Fig. 2, 3 and 4.
Fige 2- Feedback loop of fixed assets for water saving
340 THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY, CHINA
; rs ey
(we) ="
BO
Fig.3. Feedback loop of fixed assets for technology reform
Fig.4. Feedback loop of management level
4. Flow diagram of SD model for urban water using in Tianjin
Substantiate and detail the frame (feedback relation link), express various
elements and thetr relationship as various variable and their relationships
thus, the flow diagram of this model can be drawn out as shown in Fig. 5«
5. Selection and determination of parameters and TABLE functions
A series of TABLE functions, parameters and initial values in the model are
defined at last respectively according to historical statistical data, ten—
denay of objective development, planned numerical values, present situation
analysis as well as necessary judgement and through repetitive debugging.
Concrete numerical values, tables and configuration:are omitted here.
6. Analysis of experimental results
This Man-machine system was built and perfected through repetitive simi-
letion, operation and debugging of the model and through adjustment of various
parameters and input of various TABLE functions, so that we can perform
THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY. CHINA 341
Pig. -5. Flow diagram of. SD model of urban water using
342 THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY. CHINA
simlation and tracing urban water using system in laboratory, otherwise it
is impossible-to be performed in reality. Computer with the aid of its spe-
cific simulation means shows us system dynamics in different time interval,
prospect of urban water using of Tianjin in future and shows numerical values
of various water quantities concerned for the years desired.
As regards to the result of operation, we carried out goodness of fit ana~
-lysis at first. The course of history, passed has been simlated by model,
ean be used to verify whether the simulation results of the model will be in
coincidence with practical state of the real system or not, i. e. to verify
the goodness of fit of the method. According to the aboveesaid, we may check
and determine the capacity and reliability of the model to simlate system
in future in a sa¢nse, or fidelity of the model. Relationships of goodness
of fit of actual quantity of water taking and of quantity of water taking per
ten thousands yuan value of output are shown in Figs 6 and 7+
z _+. computation value
— actual value
Figs 6. Fit goodness relationship of actual water taking quantity
In addition, considering that the different proportion of development of
light and heavy industry, the different level of industry output value in
developing plan of national economy, the effect of diversion quantity of the
Yellow River, the proportion of water taking quantity for urban life and for
industry, pressure of water saving, measures to We adopted for reducing
pressure of water saving and etc., the model will carry out operation and
analysis to all these different policy choice and obtain corresponding ana~
lysis result and policy proposition. For instance, in order to reduce water
saving pressure and release unbalance between supply and demand of water
under the conditions- of present water resources, three effective ways can be
‘adopted in a definite periods a. to make an investment in water saving,
THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS SOCITY. CHINA 343
Ba --- computation value
a: — actual value
8
8
E|
gl
q
aq
aq
3
3
ie 8 8 8 @ 8 of "Te
Fig. 7 Fit goodness relationship of water taking
quantity per ten thousands yuan value of output
increase water quantity of repetitive utiliztion, improve the rate of repe-
titive utilization of water for decreasing the quatity of insustry water
taking; be to make an’ investment in processing improvement, to reform back-
ward process, which uses more water in production, into an advanced one,
which uses less water or needs no water, thus, increasing the level of ine
dustrial usage of water, as a result, decreasing the quantity of gross water
used and water intakedy c. to improve the level of scientific management of
industry water using, stop up water wasting loophole, tap the potentialities
of water saving to reach a reasonable water using,scientific management of.
water can directly decrease the quantity of industry water taking to a cer-
tain extent, relieve the pressure of water saving. ,
In short, model can privide an effective experimental method ani a field for
urban water using, especially for the policy and management of industry water
using. On these grounds administration concerned can, according to the chang-
ing of various policies, find out an effective way for solving unbalance
between supply and demand under the conditions of present water resources,
so that the limited water resources can inake maximum economic benefit
System Dynamics is a new technology of scientific management, it was intro-
duced to China recently. Management work.for urban water using just steps
344 THE 1987 INTERNATIONAL CONFERENCE OF THE SYSTEM DYNAMICS *SOCITY. CHINA
forward, a complete management system and theory are not formed yet, so there
are no ready mode and rippened method for applying System Dynamics in the
field of urban weter using, it is only an initial attempt and process of
practice. Owing to the limited knowledge of the author, there must be, de-
ficieney,and errors, Kindly give us your advice.
REFERENCES
1. Forrester, Jay W. (1961), Industrial Dynamics, Mit Dress»
2. Yang, Tongyi st al, Principle of systems, Shanghai, Translated from
Forrester, Jay W. (1968), Principle of systems.
3. Wang, Qifan (1984), Principles of Dynamic systems, Shanghaie
4. Eu,Yukui (1984), System Dynamics, Beijin.
5e Li, Baoheng, The Limits to Growth, Chengdu, Translated from Meadows,
Donella et al at The Club of Rome (1972), The Limite to Growth.
CC BY-NC-SA 4.0