cost benchmarking in the german water sector · tadeo et al. (2008) spain (andalus ia) 2001 38...
TRANSCRIPT
![Page 1: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/1.jpg)
0
PD Dr. Christian Growitsch, EWI Köln
Dr. Mark Oelmann, WIK-Consult
INFRADAY 2010, Berlin, 9 October 2010
Cost Benchmarking in the German Water Sector
Implications of an Assessment of Current German
Approaches
![Page 2: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/2.jpg)
1
Outline
• Introduction
• Literature Survey and Variables
• Data Set
• Methods
Cluster Analysis
Efficiency Analysis
• Results
• Conclusion
![Page 3: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/3.jpg)
2
Introduction
• Cost-Benchmarking in the German water sector has never been done
before, but a lot of metric benchmarking projects were carried out in
Germany.
• “Modernisation Strategy” of the Federal Government and the Bundestag:
Implementation of benchmarking-projects in each of the Bundesländer
(many of them carried out by Rödl & Partner)
• Main idea: Confidentiality of individual data and freedom to participate
leads to fruitful discussions between companies The inefficient
company in its widest sense (quality..) learns from the best performers.
• Main challenge: Differences between companies need to be explained as
much as possible for giving the best achievable feedback to utilities.
![Page 4: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/4.jpg)
3
Introduction
• Costs in a certain part of the value
chain are solely compared with
those of others without taking into
account differences in basic
conditions
• Aim of introducing established
efficiency analysis: Improve the
informative value of benchmarking
for companies by better clarifying
the deviation as the difference
between the actual value of a
company and the benchmark.
Total Costs
1.32
0.731.03
1.28 1.29 1.522.03
Value Min Q 25 Median Mean Q 75 Max
€/m³
Current Approach:
Future Approach:
Value chains already performed
Example:
![Page 5: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/5.jpg)
4
Outline
• Introduction
• Literature Survey and Variables
• Data Set
• Methods
Cluster Analysis
Efficiency Analysis
• Results
• Conclusion
![Page 6: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/6.jpg)
5
Author(s)
Region Time
Period
Number
of
Utilities
Method Measure Inputs Outputs Structural/Qualtiy Variables Effect of
Structural/Q
uality
Variables
Coelli and
Walding
(2006)
Australia 2002/03
(1995/96-
2002/03)
18 DEA total factor
productivity,
technical
change
operating
expenditure,
length of mains
number of
properties
connected,
volume of
water
delivered
percentage of non-residential
connections, percentage of
water from non-catchment
sources, average annual
rainfall, average maximum
temperature, peak of average
flow, electricity connection per
consumption
no effect
García-
Sanchez
(2006)
Spain 1999 24 DEA technical
efficiency
total staff,
network length,
treatment plants
water
delivered,
number of
connections,
chemical
analyses
population density significant
effect
Lin (2005) Peru 1996-2001 36 SFA cost
efficiency,
technical
change
wage, price of
capital
water billed,
number of
customers
accounted for water ratio,
coverage, positive rate of
chlorine tests, continuity of
service
significant
effect when
treated as
additional
outputs rather
than
environmental
variables
Picazo-
Tadeo et al.
(2008)
Spain
(Andalus
ia)
2001 38 (water
and
sewerag
e)
DEA technical
efficiency
delivery
network,
sewage
network, labor,
operational
costs
population
served, water
delivered,
treated
sewage
unaccounted for water significant
effect, but
does not
change
ranking of
sample
Benchmarking – Models and Potentials
Literature Survey and Variables International Literature Survey (Selection) (1)
![Page 7: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/7.jpg)
6
Literature Survey and Variables International Literature Survey (Selection) (2)
Author(s)
Region Time
Perio
d
Number
of
Utilities
Method Measure Inputs Outputs Structural/Qualti
y Variables
Effect of
Structural/Quality
Variables
Renzetti
and
Dupont
(2008)
Canada 1996 64 DEA technical
efficiency
expenditure on
labor,
expenditures on
materials and
energy,
distribution length
water delivered extreme
temperature, total
annual
precipitation,
dummy for
source, population
density, distance
between water
utility and highest
elevation,
residential
demand, number
of dwellings
overall exogenous
environmental variables
have significant impact on
measured input slacks and
efficiency scores
Tupper
and
Resende
(2004)
Brazil 1996-
2000
20
(water
and
sewerag
e)
DEA technical
efficiency
labor expenses,
operational costs,
other operational
costs
water produced,
treated sewage,
population served
(water), population
served (treated
sewage)
density of water
network, density
of sewage
network, index for
water loss
significant effect of density
of water network and water
loss
Zschille et
al. (2009)
Germany 2006 373 DEA technical
efficiency
revenues water meters, water
delivered to
households, water
delivered to non-
households
output density,
leak ratio,
groundwater ratio,
elevation
difference, dummy
for East Germany,
dummy for
sewage services,
dummy for other
services
significant effect of output
density, leak ratio (but
neglected in stage 3 –
controllable by
management); overall
relatively small changes in
efficiency after input
adjustment
![Page 8: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/8.jpg)
7
Energy 8%
Water 2%
Material for Water Treatment 2%
Water Abstraction Charges 3%
Other Taxes and Duties 2%
Other Operational Costs 5%
Concession Levy 10%
Interest 18% Depreciation 20%
Personnel 20%
External Services 6%
Other Material Costs 4%
The German “Hollaender”-Study (2008):
► Factor 1
Geographical conditions
► Factor 2
Habitat density and demography,
customer structure and service
area size
► Factor 3
Investment activity and cost
approach to capital costs
► Factor 4
Public funds
► Factor 5
Concession levy
► Factor 6
Water abstraction charges
► Factor 7
Compensatory payments for
agriculture
Literature Survey and Variables Additional Literature Survey
Easíly deducted in cost benchmarking
Specific conditions
![Page 9: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/9.jpg)
8
Literature Survey and Variables Variables (Selection)
Variable Unit
GROUP 4:
Area
km²
Inhabitants per m³ (area) No.
Water losses m³
Downturn in demand since 1992 %
Downturn in demand since 1998 %
Pumping Stations No.
Water produced from spring water m³
Water produced from groundwater m³
Water produced from surface water m³
No treatment of raw water m³
Simple treatment of raw water m³
Sophisticated treatment of raw water m³
Supply (adjusted for re-distribution) per tank m³
Household connections per tank No.
Variable Unit
GROUP 1:
Number of household connections
No.
Accounted water €
Transportation and distribution pipes km
Inhabitants No.
GROUP 2:
Tanks
No.
Tank capacity m³
Valves No.
Service areas No.
Height Difference M
Accounted water (excl. re-distribution) to distribution
and transportation pipes m³/m
Distribution and transportation pipes per household
connection M
GROUP 3:
Supply to re-distributors
m³
Household supply relative to accounted water
(excluding re-distribution) %
Pipe damages No.
Peak supply relative to supply of the day %
Energy consumption per produced and treated m³
of water (adjusted for re-distribution and imported
water)
kWh/m³
Chemicals per produced and treated m³ of water
(adjusted for re-distributed and imported water) €/m³
![Page 10: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/10.jpg)
9
Outline
• Introduction
• Literature Survey and Variables
• Data Set
• Methods
Cluster Analysis
Efficiency Analysis
• Results
• Conclusion
![Page 11: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/11.jpg)
10
Data Set Adjustments (1)
• Starting Point:
Total number of observations of Rödl & Partner (612 observation)
• Adjustment:
Step 1: Eliminate all observations from different years of the same
company, keeping the most current one
Step 2: Delete all observations from before 2006
Step 3: Drop all companies without any distribution network at all
Step 4: Delete all observations where crosschecks revealed
inconsistencies
• Result: 196 observations remained
![Page 12: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/12.jpg)
11
Data Set Adjustments (2)
• Revision of usable data sets:
Step 1: Adjusting data by restating 2006 data in terms of 2007 prices
(producer-price index “Water and Water Services”)
Step 2: Deduct the concession levy, the water abstraction charges,
and the compensatory payments for agriculture from total
costs to reach a maximum of comparability
• Remark:
Sample contains relatively bigger companies than the overall
German average
![Page 13: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/13.jpg)
12
• Introduction
• Literature Survey and Variables
• Data Set
• Methods
Cluster Analysis
Efficiency Analysis
• Results
• Conclusion
Outline
![Page 14: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/14.jpg)
13
Methods 1. Cluster Analysis
• Applied Hierarchical Clustering
• Average Linking Clustering identified 15 Cluster
• Reduced No. of Clusters to statistically reasonable No. of
Observations: 3 Clusters
Cluster small Cluster medium Cluster large
Lower bound 0 3,097,810 6,212,020
Upper bound 2,794,820 5,400,300 17,748,300
![Page 15: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/15.jpg)
14
Methods 2. Efficiency Analysis
a. Data Envelopment Analysis (DEA) b. Stochastic Frontier Analysis (SFA)
Output
C
Cost0
R
R„
Efficiency Frontier
P=P„
QQ„
A
B Noise effect vi<0
0 Output
Cost
Noise effect vi>0 Inefficiency
ui
• VRS
• Outlier detection: Peer-Stripping
• Cobb-Douglas, Translog &
normalized linear Functions
• Outlier corrected data set
![Page 16: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/16.jpg)
15
• Introduction
• Literature Survey and Variables
• Data Set
• Methods
Cluster Analysis
Efficiency Analysis
• Results
• Conclusion
Outline
![Page 17: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/17.jpg)
16
Results Large utilities
• Large companies: water delivery of 2.5 – 50 million m³ annually
• 59 companies
• Variables:
Total Cost (excl. concession levy, water abstraction charges,
compensatory payments for agriculture)
No. of Connections, Water delivered
• SFA Results:
Mean Standard
Deviation
Minimum Maximum
Efficiency Score 0.877 0.038 0.772 0.941
![Page 18: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/18.jpg)
17
Results Large utilities
• Rank Correlation: 0.946
![Page 19: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/19.jpg)
18
Results Medium utilities
• Medium companies: water delivery of 0.5 – 2.5 million m³ annually
• 82 companies (after outlier analysis)
• Variables:
Total Cost (excl. concession levy, water abstraction charges,
compensatory payments for agriculture)
No. of Connections, No. of Service areas
• SFA Results: Mean Standard
Deviation
Minimum Maximum
Efficiency Score 0.800 0.104 0.668 0.967
![Page 20: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/20.jpg)
19
Results Medium utilities
• Rank Correlation: 0.948
![Page 21: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/21.jpg)
20
Results Small utilities
• Small companies: water delivery below 0.5 million m³ annually
• 34 companies (after outlier analysis)
• Variables:
Total Cost (excl. concession levy, water abstraction charges,
compensatory payments for agriculture)
No. of Connections, Network (Distribution and transportation pipes)
per connection
• SFA Results:
Mean Standard
Deviation Minimum
Maximum
Efficiency Score 0.799 0.111 0.559 0.959
![Page 22: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/22.jpg)
21
Results Small utilities
• Rank Correlation: 0.790
![Page 23: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/23.jpg)
22
• Introduction
• Literature Survey and Variables
• Data Set
• Methods
Cluster Analysis
Efficiency Analysis
• Results
• Conclusion
Outline
![Page 24: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/24.jpg)
23
Conclusion
• Water distribution companies of different size do not share a common production technology
• However, No. Of connections seem to be an important cost driver for each of the three clusters
• Endogenous size cluster lead to less robust results than industry specific cluster
• Overall, average water utility efficiency seems comparably high.
![Page 25: Cost Benchmarking in the German Water Sector · Tadeo et al. (2008) Spain (Andalus ia) 2001 38 (water and sewerag e) DEA technical efficiency delivery network, sewage network, labor,](https://reader034.vdocument.in/reader034/viewer/2022050410/5f8710901b54b332e6779da2/html5/thumbnails/25.jpg)
24
Dr. Mark Oelmann
Head of Department
“Water, Sewerage & Transport”
wik-Consult GmbH
Rhöndorfer Str. 68
53604 Bad Honnef
Germany
Phone +49 2224 9225-94
Fax +49 2224 9225-66
Email: [email protected]
PD Dr. Christian Growitsch
Energiewissenschaftliches Institut
an der Universität zu Köln (EWI)
Alte Wagenfabrik
Vogelsanger Str. 321
50827 Köln
Germany
Phone +49 221-27729204
Fax +49 221 27729400
Email: