Case Analysis and Development Suggestions on China’s Water Loss
Control by Minimum Night Flow
Shuaihua Hou
1,2 a
, Feng Xu
2
, Chao Wang
2
and Tao Tao
1
1
College of Environmental Science and Engineering, Tongji University, Shanghai, China
2
Changzhou CGE Water Co., LTD., Changzhou, China
Keyword: MNF, Hourly Statistics, Short-Time Inference, Accuracy Analysis, Ma Value, Leak Determination
Abstract: Using minimum night flow (MNF) to determine leakage is an efficient method in District Metered Area
(DMA)，in which short-time inference is more useful and convenient base on the analysis with the
comparison of hourly statistics. The case analysis shows the accuracy and effectiveness of short-time
inference are mainly interfered by the precision of data recording system and the matching degree (Ma)
between the minimum flow and the diameter of metering equipment in the DMA’s inlets. Due to the current
state of affairs in China, it is not feasible to immediately upgrade data precision; therefore, it is suggested
that one could combine hourly statistics and improving Ma values in the application process of leak
determination. Moreover, it is also necessary to promote the popularization of short-time inference through
upgraded metering and recording equipment precision.
a
https://orcid.org/0000-0002-8488-3722
1 INTRODUCTION
1.1 The Basic Theory of Minimum
Night Flow
Multi-stage metering division system is recognized
as an efficient means for leakage control in
large-scale water distribution system (Perelman and
Ostfeld 2012), in which District Metered Area
(Referred to as the DMA) is the most important
basic unit (He 2018). Moreover, leakage detection is
carried out frequently within the DMA (Thornton
2003)，while the method to evaluate whether leakage
exists in the DMA depends more on MNF (Lin
2011).
In recent years, with improvement in measuring
accuracy of flow metering equipment in the water
distribution system, criteria for MNF are also
changing constantly.
Differing from the IWA definition (Al-Washali,
Sharma and Kennedy 2016), China calculates MNF
(Q
MNF
) using the customer's normal night-time water
consumption (Q
NC
), the detectable leakage (Q
DL
)
(Buchberger and Nadimpalli 2004; Li et al. 2020)
and the undetectable leakage (also termed
background leakage (Liang and Wang 2017)) (Q
UD
),
thus Q
MNF
=Q
NC
+Q
DL
+Q
UD
(Li et al. 2018).
Conventional MNF theory holds that during the
night, due to reduced human activities leading to
decreased water consumption, there exists a lowest
water consumption time at night. Moreover, under
the same water supply pipe network conditions, it is
believed that the background leakage (Q
UD
) should
be stable (Al-Washali et al. 2020) and the customer's
normal night-time water consumption (Q
NC
) should
correlate seasonally (Zheng et al. 2021)
in the same
region. Therefore, the MNF’s standard value (Q
SMNF
)
fixed for a season of a single DMA can be expressed
as “Q
SMNF
= Q
NC
+Q
UD
” without detectable leakage. If
the water supply pipe network conditions are
changed, such as the case of new leakage points,
MNF will change and the variable value is the
detectable leakage in the DMA’s water supply pipe
network which can be calculate as “Q
DL
=
Q
MNF
-Q
NC
-Q
UD
= Q
MNF
-Q
SMNF
.”
It is impossible to accurately obtain the
customer's normal night-time water consumption
and background leakage of a single DMA due to the
limitation of technical conditions and economic
benefits. However, the total MNF can be measured
Hou, S., Xu, F., Wang, C. and Tao, T.
Case Analysis and Development Suggestions on Chinaâ
˘
A
´
Zs Water Loss Control by Minimum Night Flow.
DOI: 10.5220/0011900600003536
In Proceedings of the 3rd International Symposium on Water, Ecology and Environment (ISWEE 2022), pages 61-70
ISBN: 978-989-758-639-2; ISSN: 2975-9439
Copyright
c
2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)
61
by metering equipment installed at DMA entry, and
the variable value before and after the occurrence of
a new leak can be calculated (Buchberger and
Nadimpalli 2004). Therefore, the difficulty of
judging whether a leakage occurred or not is shifted
to determining the MNF’s standard value in different
seasons at a single DMA.
1.2 Common Methods Using MNF for
Leakage Detection in China
Although most research in the industry set MNF’s
standard value based on historical monitoring
combined with confidence intervals (Buchberger and
Nadimpalli 2004; Alkasseh et al. 2013; Farah and
Shahrou 2017)
, due to the complexity and
prolongation of the process, the calculated value
based on the staged monitoring are adopted as the
MNF’s standard value more often than not.
Accounting for China's domestic application
development situation, the common methods to use
MNF for leakage determination include
Interpolation between different sequences, Hourly
statistics and Short-time inference.
Interpolation between different sequences
calculates the detectable leakage by using the
difference between the average NFM value for N
consecutive days before and after the selected
observation day (symbolized by Q
b
and Q
a
) (Li
2017). As shown in Fig. 1, if a new leak occurs on
the observed day, there are the following
corresponding quantitative and calculation
relationship:
Q
b
= Q
SMNF
, Q
a
= Q
MNF,
Qd=Q
b
+Q
DL
and Q
DL
= Q
MNF
-Q
SMNF
=Q
a
-Q
b
Although this method can effectively reduce
disturbance caused by MNF fluctuation, the
selection of N value requires long-term testing and
continuous monitoring data.
Hourly statistics is to compare the measured
MNF (Q
MNF
) with the preset MNF’s standard value
(Q
SMNF
) in the absence of detectable leakage. When
Q
MNF
>Q
SMNF
, a new leakage is determined, and the
detected leakage (Q
DL
) is the difference between
Q
MNF
and Q
SMNF
. According to the above MNF
theory, the difficulty of this method is to determine a
reasonable Q
SMNF
value, as the customer's normal
night-time water consumption and background
leakage will fluctuate with seasonal changes and
other conditions, thus Q
SMNF
will need to be updated
with actual data in any given DMA.
Figure 1: Graphical illustration of Interpolation between different sequences.
Short-time inference is an improvement on
Hourly statistics. By shortening MNF observation
interval to 2min, 5min and 15min, the disturbance of
the customer's normal night-time water consumption
is reduced as much as possible, so that the metered
Q
SMNF
is approximately equal to Q
UD
(Buchberger
and Nadimpalli 2004; Janković-Nišić et al. 2004;
Covas, Jacob and Ramos 2008). Due to the relatively
uniform characteristics and the elongated stability in
a single DMA, it is easy and quick to evaluate
background leakage. Therefore, it reduces both time
and difficulty to gain MNF’s standard value (Q
SMNF
).
Although short-time inference will simplify
Q
SMNF
’s acquisition and improve the efficiency of
leakage determination in a single DMA, its actual
application effect remains to be confirmed due to
less practical experience (Dai and Liu 2016)
especially under the current flow measurement
method in water distribution system. The following
case study on application effect will concretely
analyze the advantages and disadvantages of
short-time inference comparing to hourly statistics in
order to raise some improvement measures.
2 METHODOLOGY, DATA AND
RESULT
2.1 Design of Research Experiment
As shown in Figure 2, the internal water distribution
system of a single DMA is composed of municipal
water supply pipe network and secondary water
supply pipe network. There are two DN200 inlets of
the municipal network with water meters installed to
ISWEE 2022 - International Symposium on Water, Ecology and Environment
62
the north and south of the DMA. One DN150 inlet
water meter is installed before the water pump of the
secondary network, which can reflect the actual
water consumption since the secondary water supply
uses the pressure-superposed mode under the
condition of controlled water loss of the pump. Thus
this DMA’s Q
MNF
can be measured and calculated
through the above three water meters.
Figure 2: The DMA schematic.
With a small number of convenience stores
scattered throughout the DMA, more than 95% of
consumption unit of tap water is residential; in which
1,030 households are supplied by municipal water
supply pipe network and 1730 households by
secondary water supply pipe network. Hourly
statistics and short-time inference were used
simultaneously and MNF observation time was
selected from 00:00 to 6:00 every day in reference to
industry experience (Adlan et al. 2013; Li, Gao and
Qiao 2017)
.
Moreover, 15 minute records were
multiplied by 4 to convert to hourly record for
unified units and convenient comparison.
2.2 Data Collections
Figure 3 illustrated the MNF monitoring of
municipal water supply pipe network. 15min record
showed flow greater than zero for a long time and
hour record was greater than local empirical value
(Changzhou’s empirical value in summer was
roughly 3m
3
/h) from August 18 to August 26. Hence,
in the evening of August 25, noise leak detector was
installed. Two leaks were detected on the 26th and
repaired by the evening on the 26th. No new leak
was detected from August 27 to September 5, and 15
minute record immediately dropped to 0. However
hour record was fluctuating and in some cases, still
greater than the local empirical value; although its
extreme values were lower than before the repair,
and the general trend was lower (there will be further
analysis for the fluctuations after September 6 later).
Figure 4 illustrated the MNF monitoring of the
secondary water supply pipe network, with the same
monitoring and leak detection methods as the
municipal network. Leaks were detected on August
19 and repaired on August 20. No new leak was
detected from August 21 to September 13. 15 minute
records immediately dropped to 0, but with
intermittent fluctuation and a peak value of 4. While
the MNF’s hour records were also lower than before
repair, they were still greater than the local empirical
value in some cases with the same intermittent
fluctuation.
Case Analysis and Development Suggestions on Chinaâ
˘
A
´
Zs Water Loss Control by Minimum Night Flow
63
Figure 3: MNF records in municipal water supply pipe network.
Figure 4: MNF records in secondary water supply pipe network.
2.3 Interpretation of Results
Based on the above analysis, it is easier to determine
whether a leakage existed by 15 minute record than
hourly record because 15 minute records can better
shield the influence of the customer's normal
night-time water consumption when MNF is being
recorded. However, it is still necessary to conduct a
detailed analysis regarding the intermittent
fluctuation of 15 minute record and why the hourly
records after leak repaired were still greater than the
local empirical value.
0
2
4
6
8
10
8/18
8/19
8/20
8/21
8/22
8/23
8/24
8/25
8/26
8/27
8/28
8/29
8/30
8/31
9/1
9/2
9/3
9/4
9/5
9/6
9/7
9/8
9/9
9/10
9/11
9/12
9/13
FLpw: M
3
/H
Date:month/day
Hour Records
15 minute Records
0
1
2
3
4
5
6
7
8/18
8/19
8/20
8/21
8/22
8/23
8/24
8/25
8/26
8/27
8/28
8/29
8/30
8/31
9/1
9/2
9/3
9/4
9/5
9/6
9/7
9/8
9/9
9/10
9/11
9/12
9/13
Flow: M
3
/H
Date:month/day
Hour Records
15 minute Records
Leak repaired
Leak repaired
ISWEE 2022 - International Symposium on Water, Ecology and Environment
64
3 DISCUSSION
3.1 Accuracy Analysis on 15min
Record
3.1.1 Accuracy Impact of Data Recording
Equipment
The unit measurement of metering equipment for
water supply pipe networks can reach per liter or
higher precision. However, in present day China, the
data recording equipment precision for metering is
in cubic meters in most water supply systems, which
means the collected flow value less than 1 m³ cannot
be recorded and will show a rounded value. As a
result, when the flow value collected by metering
equipment is less than 0.5 m³ per 15 minutes, the
equipment will record the data at value 0 m³; if the
flow value is greater than 0.5 m³ but less than 1 m³
per 15 minutes, the equipment will record the data at
value 1 m³. The MNF’s 15 minutes records in Fig.3
and Fig.4 were the minimum value collected per 15
minutes every night, of which the time interval was
between 20-25 hours, so the customer's normal
night-time water consumption was close to zero. In
addition, since the pipes in this DMA was relatively
new, there was less background leakage. Therefore,
the MNF’s 15 minute records after leaks were
repaired should be less than 1 m³/15min and might
float around 0.5 m³/15min which meant the value
alternately appeared to be greater than 0.5 and less
than 0.5 due to the progressive counting. On account
of accuracy, the value collected above couldalternate
between 0 m³ and 1 m³ after being fed back to the
recording equipment, and alternate between 0 m³
and 4 m³ after being converted to hourly record as
shown in Fig.3 and Fig.4.
The same situation could be verified by the
continuous MNF’s 15 minute records of the
secondary water supply pipe network for a period of
time in a day. As shown in Fig.5, the solid blue line
represent the change trend of the actual 15 minute
records while the dotted orange line represent the
converted hour record which showed greater value
fluctuation.
Figure 5: 15 minute records and calculated value from
0:00 to 6:00 on 21 August in secondary water supply pipe
network.
Figure 6: 15 minute records from 0:00 to 6:00 on 22
August in secondary water supply pipe network.
In addition, low precision of recording
equipment destroys the smoothness of the data curve
which was represented in the form of broken lines,
as shown in Fig.6.
3.1.2 The
Ratio Impact Between
Instantaneous Flow and Meters’
Diameter
It could be seen from Fig.3 and Fig.4 that the
alternation frequency of 0 and 4 in secondary water
supply pipe network was significantly more than in
the municipal water supply pipe network after leaks
were repaired. For example, after the leak was
repaired in the municipal water supply pipe network
and the 15 minute record dropped to 0 m³ on August
26, there was an interval of 10 days until the next
fluctuation occurred on September 6. However, in
0
5
10
15
20
25
30
00:00
00:30
01:00
01:30
02:00
02:30
03:00
03:30
04:00
04:30
05:00
05:30
06:00
Water volume m
3
Time: Hou-Minute
8.21-15 minute Records
8.21-Hour Records
5
4
3333
1
2
11
2
1
2
111
0
11
0
11
22
3
0
1
2
3
4
5
6
00:00
00:30
01:00
01:30
02:00
02:30
03:00
03:30
04:00
04:30
05:00
05:30
06:00
Water volume m3
Time: Hour-Minute
8.22-15…
Case Analysis and Development Suggestions on Chinaâ
˘
A
´
Zs Water Loss Control by Minimum Night Flow
65
secondary water supply pipe network, after the
repair on August 20, the next fluctuation occurred on
August 24, only 4 days apart. Moreover, as many as
five fluctuations occurred between August 20 to
September 13, and the shortest interval was less than
one day.
The reason for such difference could be
attributed to the deviation of the matching degree
(symbolized by Ma) between the minimum flow and
the diameter (Guo and Lin 2019; Zheng 2016) of
metering equipment in the DMA’s inlets. For
research purposes, the Ma value could be translated
into the ratio between the number of households and
the diameter of DMA pipe network inlets, as shown
in Table 1.
The calculation of Ma value in Table 1 was based
on the premise of meeting the basic water needs in
the DMA. It could be seen from Table 1 that the Ma
value of municipal water supply pipe network was
much smaller than that of secondary supply pipe
network, indicating that the redundancy of inlet
diameter of municipal supply pipe network was
greater because it must be responsible for outdoor
fire hydrant water supply. However, as the main
water consumption within DMA came from
households, a greater Ma value showed larger
domestic water consumption which could be
evidenced by the fact that the continuous 15 minute
records of the secondary supply over a period of
time were generally greater than those of the
municipal supply as shown in Fig.7.
Although the measurement accuracy of
mechanical water meters have been greatly
improved with technological progress, their
measurement accuracy for small flow rate will
decrease correlating with the increase of inlet
diameter to flow rate ratio (Zhang et al. 2020; Yuan
Da and Chen 2019), which can also be considered
that the lower Ma value leads to increase in MNF
measurement error. The specific reflection in this
research was that the two DN200 water meters with
lower Ma value at the inlet of municipal water
supply pipe network might miss meter or meter less
during low flow rate, which leaded to increased
recording of flow lower than 0.5m
3
/15min and cost
long time to accumulate to the records greater than
Table 1: Calculation of the Ma value.
Supply pipe network Households(H) Pipe inlet Diameter (D) Number(N) Ma=H/(D×N)
Municipal water 1030 DN200 2 2.575
Secondary water 1730 DN150 1 11.533
Figure 7: 15 minute frequency pipeline flow monitoring.
0
2
4
6
8
10
12
14
16
18
20
8/19
8/19
8/20
8/20
8/21
8/21
8/22
8/22
8/23
8/23
8/24
8/24
8/25
8/25
8/26
8/26
8/27
8/27
8/28
8/28
8/29
8/29
Water volume :m
3
Date:month/day
Municipal water supply Secondary water supply
ISWEE 2022 - International Symposium on Water, Ecology and Environment
66
0.5m
3
/15min. Meanwhile, the only one DN150 water
meter with much higher Ma value at the inlet of
secondary water supply pipe network could meter
low flow rate precisely, which made shorter time to
accumulate to the case greater than 0.5m
3
/15min. As
a result, the situation shown in Fig.3 and Fig.4
appeared, Fig.7 also provided the same evidence as
more alternations between 0 and peak value of
MNF’s 15 minute records appeared in the secondary
water supply pipe network. Considering that most
water meters with diameters of 200 and below are
mechanical, this measurement error should be paid
attention to.
3.1.3 The Mutation Impact Caused by Flow
Change
The above discussion explained why there were
intermittent fluctuations of MNF records in DMA
water supply network after leaks were repaired, but
it doesn’t explain the continuous peak value of the
15 minute records from September 7 to September
10 shown in Fig.3; other operations involving the
flow changes need to be further considered (Arregui
et al. 2006).
In order to further verify the low Ma value’s
impact on the measurement in municipal water
supply pipe network, the DMA’s north inlet was shut
off on September 6, all municipal water flowed in
from the south inlet, causing the Ma value to double.
As shown in Fig.3, both 15 minute and hour records
of MNF increased significantly on September 7 and
lasted for about three days, after which all records
returned to a normal steady state. This test indicated
that flow mutation had a greater and longer impact
than Ma value on measurement accuracy.
3.2 Influencing Factors and
Application Status of MNF
Evaluation Mechanism
Based on the above discussion, it is easy to conclude
the main factors influencing MNF's application in
leakage assessment:
(1) Interference between the measuring frequency
and customers’ normal night-time water
consumption.
Relevant experiments show that the measurement
frequency of pipeline flow is inversely proportional
to the interference from the customer's normal
night-time water consumption, that is, higher the
measurement frequency, lesser the interference,
sometimes even negligible.
(2) The measuring precision of metering equipment
and recording precision of data equipment.
These two kinds of precision can limit the previous
factor as it is impossible to accurately determine the
leak existence by short-time inference with higher
measuring frequency if any equipment’s precision
fell short. This factor is the biggest shortcoming in
promoting widespread short-time inference leak
detection. For water supply enterprises in most
Chinese cities, the smallest measuring unit of
metering equipment is in cubic meters and the
recording precision of data equipment is per hour,
rarely the per 15 minutes needed.
Taking the MNF record of the municipal water
supply network in this DMA experiment as an
example, limited to the low recording precision of
data equipment and small Ma value of pipe network
inlet, 0 value also appeared in the 15 minute records
at the leak judgment stage which affected the
accuracy of leak determination, though the value
lasted for a short time and were different from that
after leak repaired. Therefore, if 15 minute records
changed to per 5 minutes, it can be speculated that
more 0 value records will continue to appear and
interfere with the accuracy of leak detection.
(3) Pipeline state and background leakage.
This factor’s interference is relatively easy to be
evaluated compared with the above two. Benefited
from the popularization and widespread application
of geographic Information System (GIS) in China’s
water distribution system as well as large-scale
pipeline more standardized renewal and construction,
the basic physical information of supply pipelines
can be gained easily and the source ledger can be
traced. Based on above, the water supply networks
with the same pipeline attributes in a DMA can be
classified and recorded in order to a relatively
accurate assessment of the background leakage,
reducing the influence due to differences in pipeline
conditions.
(4) Compound judgment criteria for leakage
determination.
Case Analysis and Development Suggestions on Chinaâ
˘
A
´
Zs Water Loss Control by Minimum Night Flow
67
Table 2: The MNF of leak repaired DMAs.
DMA's
Name
MNF
(Hour
Record)
(m
3
/hour)
Households
（H)
Consumption
per Household
（L/hour）
Pipe inlet
Diameter(D)
Number
Ma=
H/(D×N)
DMA-1 0 215 0 DN150 2 0.72
DMA-2 0 164 0 DN150 2 0.55
DMA-3 1 1910 0.524 DN200 2 4.78
DMA-4 1 377 2.653 DN100 1 3.77
DMA-5 3 1110 2.703 DN150 1 7.40
DMA-6 4 1413 2.831 DN200 2 3.53
DMA-7 5 1740 2.874 DN200 1 8.70
DMA-8 5 1679 2.978 DN200 2 4.20
DMA-9 5 1841 2.716 DN200 2 4.60
DMA-10 1 534 1.873 DN150 2 1.78
In the actual application process of MNF for
leakage determination, hourly statistics may be more
tolerant to low precision metering equipment and
low Ma value. Taking 10 DMAs with hourly
statistics for leak determination as an example in
Changzhou, China, in which most buildings are
residential and pipeline state are almost same with
no leak could be detected, the monitoring and
statistical data for July-August were shown in Table
2 with an accuracy of 1 m
3
.
The following content can be drawn from the
table:
① Based on the stability data from DMA-3 to
DMA-9, it could be calculated that the customer's
normal night-time water consumption in the same
type DMA was about 2.6L/Hour to 2.9L/Hour in
summer for MNF.
② The reason why the MNF values in DMA-1
and DMA-2 were zero could be speculated that the
Ma value is too low to precisely meter the small low,
and the same reason for low MNF value in DMA-10.
Through investigation, the reason for DMA-3’s low
MNF was determined to be caused by low
occupancy rate in the area.
③ The acceptable MNF can be calculated by
hourly statistics when Ma value is greater than 4,
that further proof hourly statistics’ tolerance is
stronger than short-time inference’s for low Ma.
4 CONCLUSION AND
SUGGESTION
In the actual application process of MNF for leakage
determination, short-time inference can provide
information quickly and concisely, but need higher
precision equipment to do so. As equipment get
continuously renewed and upgraded, leak detection
using short-time inference will become more
widespread.
Short-time inference is the trend of MNF for
leakage determination, upgrading China’s water
supply enterprises’ equipment will take a long
process, and the equipment should match with the
leak control requirements and the improvement of
leak detecting technology for water distribution in
China. Combined with the existing facility
conditions, there are roughly three following
suggestions for leak control at the current stage:
(1) Through sorting out the impeccable GIS
information of water distribution system, accurate
pipeline attributes can be gained so that reasonable
background leakage can be assessed.
(2) With complete DMAs’ user and consuming
information, the matching degree (Ma) between the
minimum flow and the diameter of metering
equipment in the DMA’s inlets can be accurately
calculated. On this basis, for residential-based
DMAs, Ma value should be improved, and
small-caliber non-residential meters shouldn’t be
ISWEE 2022 - International Symposium on Water, Ecology and Environment
68
greater than 15%. Industrial and commercial water
meter with diameter over DN40 are not
recommended to be included in residential-based
DMA.
(3) For new DMA or newly installed water meters,
high-precision metering equipment and
high-precision data recording system should be
applied immediately in order to promote the
popularization of MNF’s short-time inference for
more convenient leak determination.
DATA AVAILABILITY
STATEMENT
The research presented in this paper was part of the
daily work content of pipeline leakage control in
Changzhou CGE Water Co., LTD. All the data came
from daily monitoring work which all authors
participated in.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the
authors.
REFERENCES
Adlan, Mohd Nordin, Jaber M. A. Alkasseh, Hj. Ismail
Abustan and Abu Bakar Mohamad Hanif. 2013.
"Identifying the appropriate time band to determine the
minimum night flow: a case study in Kinta Valley,
Malaysia." WATER SCIENCE AND TECHNOLOGY-
WATER SUPPLY 13 (2):328-336. doi:
10.2166/ws.2013.026.
Al-Washali, T., S. Sharma, and M. Kennedy. 2016.
"Methods of Assessment of Water Losses in Water
Supply Systems: a Review." Water Resources
Management 30 (14):4985-5001. doi:
10.1007/s11269-016-1503-7.
AL-Washali, Taha, Saroj K. Sharma, Robert Lupoja,
Fadhl AL-Nozaily, Mansour Haidera and Maria D.
Kennedy. 2020. "Assessment of water losses in
distribution networks: Methods, applications,
uncertainties, and implications in intermittent supply."
Resources, Conservation & Recycling 152 (104515).
doi: 10.1016/j.resconrec.2019.104515.
Alkasseh, Jaber M. A., Mohd Nordin Adlan, Ismail
Abustan, Hamidi Abdul Aziz and Abu Bakar
Mohamad Hanif. 2013. "Applying Minimum Night
Flow to Estimate Water Loss Using Statistical
Modeling: A Case Study in Kinta Valley, Malaysia."
Water Resources Management 27 (5):1439-1455. doi:
10.1007/s11269-012-0247-2.
Arregui, Francisco, Enrique Cabrera, Ricardo Cobacho
and J. García-Serra. 2006. "Reducing Apparent Losses
Caused By Meters Inaccuracies." Water Practice &
Technology 1 (4):wpt2006093. doi:
10.2166/wpt.2006.093.
Buchberger, Steven G., and Gayatri Nadimpalli. 2004.
"Leak Estimation in Water Distribution Systems by
Statistical Analysis of Flow Readings." Journal of
Water Resources Planning and Management 130
(4):321-329. doi:
10.1061/(ASCE)0733-9496(2004)130:4(321).
Covas Dídia I. C., Ana Cláudia Jacob and Helena M.
Ramos. 2008. "Water losses' assessment in an urban
water network." Water Practice & Technology 3
(3):1-9. doi: 10.2166/wpt.2008.061.
Dai Huan-fang and Liu Shu-ming. 2016. "Primary
investigation on performance indicators for leakage
control in water distribution systems." Water
＆
Wastewater Engineering (S1):258-261. doi:
10.13789/j.cnki.wwe1964.2016.0391.
Farah, Elias and Isam Shahrour. 2017. "Leakage Detection
Using Smart Water System: Combination of Water
Balance and Automated Minimum Night Flow."
Water Resources Management 31 (15):4821-4833. doi:
10.1007/s11269-017-1780-9.
Guo Jin-Kui and Lin Yi. 2019. "Research on reasonable
selection of water meter." Water & Wastewater
Engineering (S1):159-163+165.
doi:10.13789/j.cnki.wwe1964.2019.S1.42.
He, Jian-rong. 2018. "Exploration and practice of grid
management mode of urban water supply." City and
Town Water Supply (01):78-81. doi:
10.14143/j.cnki.czgs.2018.01.015.
Janković-Nišić, Bojana, Čedo Maksimović, David Butler
and Nigel J. D. Graham. 2004. "Use of Flow Meters
for Managing Water Supply Networks." Journal of
Water Resources Planning and Management 130
(2):171-179. doi:
10.1061/(ASCE)0733-9496(2004)130:2(171).
Li, Hao. 2017. "Research on minimum night flow method
of DMA leakage prediction." Proceedings of the 12th
International Conference on China Urban Water
Development EXPO of New Technologies and
Facilities.
Li Jian-yu, Gao Jin-liang and Qiao Yi-chao. 2017.
"Economic benefit model of reducing valve by
minimum night flow." Journal of Harbin Institute of
Technology 49 (8):55-59.
Li Lan, Wu Shan, Kou Xiao-xia and Song Ling-shou.
2018. "Research progress of minimum night flow
based on District Metering Area." Water &
Wastewater Engineering (06):135-141. doi:
10.13789/j.cnki.wwe1964.2018.0183.
Li Pu-jian, Gao Jin-liang, Zhang Huai-yu and Zhang
Tian-tian 2020. "Summary of the frontier technology
and experience of leakage control in urban water
supply networks." Water & Wastewater Engineering
(06):52-57+64. doi:
Case Analysis and Development Suggestions on Chinaâ
˘
A
´
Zs Water Loss Control by Minimum Night Flow
69
10.13789/j.cnki.wwe1964.2020.06.010.
Liang Jin-ling, and Wang Qi 2017. “Pipeline leakage
control strategy based on active pressure management."
Water & Wastewater Engineering (05):111-113. doi:
10.13789/j.cnki.wwe1964.2017.0186.
Lin, Chao-yang. 2011. "The minimum night flow used to
control the non- revenue water in large water
distribution system." Water & Wastewater
Engineering (09):101-104. doi:
10.13789/j.cnki.wwe1964.2011.09.040.
Perelman, Lina, and Avi Ostfeld. 2012.
"Water-Distribution Systems Simplifications through
Clustering." Journal of Water Resources Planning
and Management 138 (3):218-229. doi:
doi:10.1061/(ASCE)WR.1943-5452.0000173.
Thornton, Julian. 2003. "Managing Leakage by Managing
Pressure: A Practical Approach." Water 5 (5):43-44.
Yuan Jun, Da Yue-wu and Chen Yuan. 2019. "Analysis of
problems in daily operation of flowmeter for DMA."
Water & Wastewater Engineering 45 (02):117-121. doi:
10.13789/j.cnki.wwe1964.2019.02.022.
Zhang Jia-xin, He Xin, Zhai Yuan-xin, Qiang Zhi-min and
Xu Qiang. 2020. "Data quality assessment and error
data identification methods for DMAs." Water &
Wastewater Engineering 46 (03):134-138. doi:
10.13789/j.cnki.wwe1964.2020.03.025.
Zheng Dong, Ni Xiao-dong, Liu Chao, Jiang Yong-wei
and Xu Zhe 2021. "Discussion on monitoring of water
loss in active residential areas at night." City and
Town Water Supply (02):97-99. doi:
10.14143/j.cnki.czgs.2021.02.019.
Zheng, Wei. 2016. "Selection and application of
large-diameter water meter." Water Technology 10
(06):34-38.
ISWEE 2022 - International Symposium on Water, Ecology and Environment
70
