- Email: [email protected]
A combined model to assess technical and economic consequences of changing conditions and management options for wastewater utilities
Journal of Environmental Management 207 (2018) 51e59
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Research article
A combined model to assess technical and economic consequences of changing conditions and management options for wastewater utilities €nckner b Mathias Giessler a, b, *, Jens Tra a b
Emschergenossenschaft/Lippeverband, 45128 Essen, Germany University of Rostock, Department for Water Management, 18059 Rostock, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 April 2017 Received in revised form 22 October 2017 Accepted 6 November 2017 Available online 12 December 2017
The paper presents a simplified model that quantifies economic and technical consequences of changing conditions in wastewater systems on utility level. It has been developed based on data from stakeholders and ministries, collected by a survey that determined resulting effects and adapted measures. The model comprises all substantial cost relevant assets and activities of a typical German wastewater utility. It consists of three modules: i) Sewer for describing the state development of sewer systems, ii) WWTP for process parameter consideration of waste water treatment plants (WWTP) and iii) Cost Accounting for calculation of expenses in the cost categories and resulting charges. Validity and accuracy of this model was verified by using historical data from an exemplary wastewater utility. Calculated process as well as economic parameters shows a high accuracy compared to measured parameters and given expenses. Thus, the model is proposed to support strategic, process oriented decision making on utility level. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Process-based model Economic quantification Wastewater utilities Changing conditions
1. Introduction In contrast to the global trend many regions in Europe suffer from a decrease and aging of population, often described as “demographic change”. In parallel, legal, technical and economic conditions are changing with integrally acting consequences for utilities. Assessment of the effects and development of appropriate adaptation measures demand for approaches which are able to describe the causal and complex relationships between changing conditions, technical system and economic effects. The shrinking population, high fix costs and long depreciation periods create economic but also technical challenges. To adapt timely and sustainable, it is important to quantitatively predict technical and economic effects of those changes and to evaluate the effectiveness of subsequent measures. There exist a lot of valuable programs for simulating processes of WWTP, sewer system as well the aquatic environment with which the effects of changing conditions can be considered. These programs are generally focused on certain systems and questions (like hydrodynamic sewer models, rehabilitation models, activated sludge models etc.). Generally, they are complex and require a high data density and thorough
* Corresponding author. Emschergenossenschaft/Lippeverband, 45128 Essen, Germany. E-mail address: [email protected] (M. Giessler). https://doi.org/10.1016/j.jenvman.2017.11.016 0301-4797/© 2017 Elsevier Ltd. All rights reserved.
parameterization. Even cost management models for efficiency analysis and decision support tools provide no holistic approach to reflect the cost of utility administration, sewer and treatment plants. Different approaches for both sectors are presented in literature. For sewer maintenance cost, a dynamic systems model developed by Rehan et al. (2014) to support the development of financially sustainable management strategies. Hernandez-Sancho et al. (2011) developed a cost modeling methodology for wastewater treatment processes based on statistical information by aiming at a better understanding of the cost structure of wastewater treatment processes. Also based on data envelopment analysis (DEA) Castellet and Molinios-Senante (2016) developed a nonradial approach by integrating technical, economic and environmental issues for efficiency evaluation of WWTPs. Ruiz-Rosa et al. (2016) designed and adapted a cost management model to calculate the costs of each activity involved in wastewater treatment and reuse processes by combining DEA and Life Cycle Assessment (LCA) indicators. On utility level, models exist to estimate the performance and efficiency. One example is the DEA model by Ferreira da Cruz et al. (2013) which measures separately the efficiency of each service. In combination with other indicators, this model can be used for prioritizing efforts to improve overall efficiency. An intrautility performance management model was proposed by Haider et al. (2016) as a decision tool for sustainability assessment of small and medium sized water utilities.
52
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
An integrated technical and economic planning, especially if changing boundary conditions are regarded, a process-oriented approach is required. This has to combine models, which realistically can describe the causal relationships within the actual technical system with an economic assessment tool, reflecting the cost structure of the respective utility. A rather pragmatic approach for such a combined assessment has been proposed by the DWA-Workgroup WI-1.3 and illustrated for a defined “standard water utility”. This method based on a simplified simulation that uses corporate benchmarking indicators. For selected trends, the economic effects of and their superposition can be derived. The model serves to generalize economic effects on changing conditions in the water sector, but depends on the availability of utility specific benchmarking data. The relationships between changing boundary conditions, infrastructure, and economic effects are rather conceptual. Complex interactions or e.g. changes in organizational structures and far going changes of the €nckner et al., technical system cannot be adequately reflected (Tra 2014). To achieve a balance between limited data availability and expressiveness a process oriented but still sufficiently simple model was developed. This paper presents The structure of the simplified process-based and economic model focused on publicly available infrastructure data, WWTPs data and expenses of the utilities. The validation of the model based on historical data of an exemplary utility in Mecklenburg-Vorpommern, Germany. An exemplary quantitative forecast of technical and economic effects till 2050 due to assumed future changes. The following sections describe the model development and its structure as well as its components (modules) and input parameters. Based on data from an exemplary utility, the model is tested, validated and the results and accuracy are discussed. 2. Model development and structure The model is programmed in MATLAB®, a commercial numerical computing environment provided by MathWorks®. It comprises substantial cost relevant assets and activities of a typical German wastewater utility limited to the core tasks wastewater disposal and treatment. It is subdivided into three modules (see Fig. 1). The module Sewer is a simplified asset management tool of the sewer system, modeling the condition as function of aging and defined activities for repair/renovation/renewal. The module WWTP is a process oriented tool calculating inflow and outflow parameters of the actual specific WWTP as well as cost relevant technical parameters. All necessary parameters from the module Sewer and WWTP are used in economic assessment module Cost Accounting. 2.1. Data set In order to specifically address the different structural, operational and administrative conditions of a specific utility, the model demands a rather comprehensive data set. In the phase of model development, data collection was rather challenging due to the data's sensitivity and the required time-consuming preparation. Therefore, previously collected data provided by statistical institutes and ministries were accessed. Table 1 gives an overview of the collected data and their origins. The data of statistical institutes and ministries were taken at municipality level. The service area of a utility often differs from the official administrative borders. This required structured and reproducible approaches to compare process and assign the actual
service area to available data. For the case study, the data and their interrelations were double-checked for accuracy and plausibility by the stakeholder. In addition, individual surveys were conducted to obtain stakeholder data.
2.2. Module sewer The module Sewer describes the development of sewage systems state due to aging and subsequent rehabilitations measures. Input parameters for the module sewer are i) the classification of the sewer network into age groups (<10a, >10a, >20a, >30a, >40a, >50a) and ii) condition states (CS) according to DWA-M149-3 (2015) from CS 5 (best condition, meaning new sewers with no defects) to CS 0 (worst condition, meaning old sewers at the end of their life cycle with many or great defects). Calculation of age groups, determines the end of depreciation from network parts. Due to structural deterioration, the sewage system does not develop linearly, rehabilitation may be necessary at earlier stages. Therefore, CS are estimated to determine which parts of the sewer network need rehabilitation. The sewer system is rehabilitated according to age and condition, beginning with the oldest age group and worst CS and ending with the younger age groups and better CS. The rehabilitations and developments are calculated in annual time steps for the cost calculation while transitions between age groups occur in ten years intervals. For transition of CS, the cohort survival function from Herz (1996) is used. This function, called the Herz-distribution, calculates the annual percentage of sewer parts which stay in the same CS (cohort). The parameters of the survival functions, including the aging vector, the transition vector and the resistance vector, were calibrated according to Jansen (2007), who describes the calibration for the program AQUA-WertMin. The parameters can be adjusted to the specific sewer system. This approach is also integrated in other software products like DynaStrat and KANEW-Z, which are mainly proposed by German consulting offices (Kley and Caradot, 2013). In context of long term strategies and budget requirements, the cohort survival approach is very useful and is simple to compute (Ana and Bauwens, 2010; Kley and Caradot, 2013). However, the Herz-functions seem to overestimate the actual survival rates and remaining life expectancies (Le Gat, 2008). Beside the statistical deterioration approach from Herz, there are countless different deterioration models which could alternatively implemented. Several review papers give a good overview of the available models. Yang (2004) categorized the model types into three classes: physical models, artificial intelligence based models (e.g neural network, Fuzzy set theory), and statistical models. Rajani and Kleiner (2001) give an overview and description of physical models. They subdivided them into probabilistic and deterministic models. The companion paper of Kleiner and Rajani (2001) focuses on statistical models and divide this class into deterministic models with a subdivision in time exponential models and time linear models as well as probabilistic models with the subdivision in multi-variate models (proportional hazards models, accelerated lifetime models, time-dependent poisson models) and singlevariate group-processing models (cohort survival model, Bayesian diagnostic model, semi-markov model, break clustering). Ana and Bauwens (2010) subdivide the statistical models generally into two groups: pipe group models to predict the condition of a sewer group (cohorts) and pipe level models to simulate each single pipe. For all groups different models were presented and described in the paper. Because of the data set in this study, only statistical sewer group models can be used. Models to simulate individual pipes need a more detailed data set.
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
53
Fig. 1. Modular schema of the created model.
Table 1 Collected data with their origin. Data
Source
Type of survey
Dataset
Sewer length Sewer age Kind of sewage disposal (centralized/ decentralized) Population equivalent Sewage (type, amount) Population Areas according to land utilization Number of buildings and apartments
Statistical State Office MV (StatA MV)
Federal Statistics
structural sewer data
Design capacity Process Technology Construction year and capacity Nominal load In- and Outflow parameters (Q,COD,N,P) Permit values (COD,N,P)
State Office for Environment, Nature Conservation and Geology MV (LUNG)
Survey of the LUNG base on Wastewater Levy Act (x 13 (1) AbwAG)
P Elimination Sludge treatment/disposal CoSubstrate Energy consumption/production Charge Annual report (balance sheet, income statement) Personnell structure Levying/Reversal of contribution Funding/release subsidies Asset values
Wastewater utility
Own survey
structural data municipality
Micro census
2.3. Module WWTP Module WWTP serves to simulate effects of changing conditions in wastewater treatment plants. Running a dynamic process model as the activated sludge models (ASM) 1 to 3 (Henze et al., 2000) is from the perspectives of effort for setting up the model, the data availability and desired time resolution not advisable. Instead a numerical mass-flow model, based on annual input data was developed. Based on quantified mass flows, operational costs calculated.
Object data WWTP
Economy data
Module WWTP simulates sludge production (primary sludge, activated sludge), demand of precipitants and flocculants, sludge treatment and disposal, energy production of digester gas and electricity consumption of the plant. Therefore, conventional WWTP- including nutrient removal and the option of anaerobic sludge stabilization as well as WWTP with aerobic sludge stabilization and the option of anaerobic pretreatment for special inflow conditions are taken into account. Based on ATV-DVWK A 131 (2000) Annex 2, the activated sludge system is calculated basically by combining of chemical oxygen demand (COD), nitrogen (N)
54
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
and phosphorous (P) mass balances with kinetic functions of growth and decay. So far, average standard values of the guideline are used for calculation but can be replaced if more detailed information is available. Sludge production in rotational biological contactor (RBC) plants and trickling filters is calculated according to ATV - DVWK - A 281 (2001). For primary settlers with one hour detention time, ATV-DVWK A 131 (2000) proposes an efficiency COD 35%, total suspend solids (TSS) 60%, N 10% and P 10%. For pond treatment systems, the cleaning performance for organic matter of a tailing pond as a primary clarifier stage is according to DWA-A 201 (2005) assumed with 30e50%. WTTP's with a design capacity 10,000 population equivalent (p.e.) are in Germany often designed for simultaneous aerobic sludge stabilization without primary clarification and as default described in the model accordingly. Gas yield and, subsequently, the generated energy are calculated according to Gretzschel et al. (2014). This calculation depends on the sludge conditions and the presence of co-substrates. The reduction of sludge by anaerobic treatment is taken into account with a factor by Imhoff and Imhoff (2007). Estimation of sludge mass to be disposed depends on dewaterability. Values for this parameter in conjunction with sludge as well as process technology (aerobic or anaerobic) and procedural steps (conditioning, dewatering, drying) like in the presented state-variables selected for their relevance to WWTP sludge management according to Silva et al. (2016), were taken from DWA-M-366 (2000). In comparison to that model, module WWTP has no evaluation of the sludge management performance. Energy consumption of treatment processes are determined by p. e. specific consumption depending on process technology. For plants which produce electricity through sludge digestion, the model assumes that electricity production covers the plants own energy demand first, leading an according reduced net consumption. If due to special conditions more energy is produced than required the surplus is fed into the grid. The calculation of pollution units for wastewater discharge fees is based on German wastewater Levy Act x3 (AbwAG, 2005). COD, N and P are regarded, while heavy metals are neglected due to missing data. 2.4. Module cost accounting The module Cost Accounting is based on the State Municipal Charges Act (KAG M-V, 2005) and results in the cost structure of the utility as well as derived charges to the wastewater producers. Beside the above described consideration of the actual technical system and technology, it takes the utility specific organizational and financial structures into account. Results from the modules WWTP and Sewer converge in this calculation, and are accounted and allocated to the respective cost category. All expenses are summarized and represented by the cost structure. Charge-reducing liquidation of contributions and subsidies are, if given, counted and updated. If sewers are newly connected, an actual contribution rate is determined in accordance with the KAG M-V x 7. Considering the investments of WWTP and sewers, resulting interest through debt financing and depreciation, future assets and liabilities are simulated. All information ends in the calculation of charges under condition of cost coverage. The rules for cost accounting and charging are differently set, even by the federal states within Germany. Application of the model in other regions requires according adaptions. Table 2 shows how chargeable costs are composed and which costs of the cost category have been considered by technical parameters in the respective cost centers. All costs in the model are subject to inflationary influences.
Unless otherwise indicated, the rise of consumer prices was taken as inflation. With regard to contrasting or intensified trends in price developments compared to the consumer prices (StatBA, 2017) in the past, separated inflation rates for electricity (BDEW, 2015), grid feed (BMWi, 2016), sludge disposal (requested by the utilities) and sewer cleaning (requested by the utilities) are used. For the construction of sewage systems (sewer, WWTP) the construction index for civil engineering (StatBA, 2016) has been set. Generally the personnel cost in the model results on the number of employees and wage developments. This cost category is divided into administration, WWTP and sewer. Due to different structures in the organization, specific ratios for the detection of the required number of employees in the administration are too inaccurate. Furthermore, certain processes, such as sewer rehabilitation and cleaning or sludge disposal can also be outsourced. In the model, these costs are in the cost category purchased services, so they not directly influence the personnel costs. Prospected personnel strength at the end of the simulation period is estimated and must be indicated. For WWTP-staff, the workload for maintenance and operation is taken into account according to the state ordinance on self-monitoring of sewage plants and sewage discharges. In case of a change in the WWTP size class or process technology, the increase or decrease in expenditure by man-hours is counted to personnel costs. Electricity costs are calculated by electricity consumption, electricity price considering the cost increase. The costs for precipitants are based on consumed mass, price and inflation rate for consumer prices. Incomes from surplus of electrical energy which are fed into grid are invoiced with investments. The remaining share of operating costs has been summarized and due to this share is mainly composed of costs from sewer network operation it was assumed that these costs are changing proportionally to dry weather flow. Sludge disposal costs are accounted as purchased services. The costs depend strongly on the disposal way (agricultural or incineration). Prices are defined per cubic meter to be disposed and a derived specific price increasing rate. Costs for sewer network repairs results from the CS of the sewer network in module Sewer. Only CS 0 to 2 are declared as need for repair. For each of these three CS, a specific price is determined per meter. The worse sewer condition the higher is expense of repair cost. The remaining expenses in cost category purchased services are summarized and proportionally to connected p. e. The calculation of the cost categories depreciation and interest is constructed in a functionally same manner. The total amount of depreciation and interest results from the shares of WWTP and sewer as well as the summary of other tangible assets under others. Depreciation is distinguished between imputed depreciation and depreciation of acquisition and production costs. The kind of depreciation can be selected by the user. According to KAG M-V x 6 (2a), linear depreciation is applied. In addition to the acquisition and production costs, assets to be reinvested are also included in the calculation. The asset age is taken from the modules WWTP and Sewer and the end-of-service life is determined. In the case of reinvestment (depending on the strategy), the replacement value is calculated at the respective point in time. From this investment value, interest and depreciation are calculated and added to the respective cost types. The remaining asset-values, which are not part of WWTP or sewer network and not given in the data, are summarized in one position. For the model, it was assumed that this position mainly contains administrative assets (buildings, vehicles and others). By reason of the dominant share of buildings the asset-value, the depreciation period of this position was set equal to that of buildings with 50 years.
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
55
Table 2 Considered parameters of cost categories within the cost centers. Cost category
Administration
WWTP
Sewer network
Personnel Operating Purchased services Depreciation Interest Other expenses
Personnel costs for administration
Personnel costs for WWTP Electricity Precipitant Sludge disposal WWTP WWTP Wastewater levy
Personnel costs for sewer network
Buildings and others Buildings and others
e
P Chargeable costs Reversal of connection charges
e
Reversal of subsides
¼
Costs set equal to expenses
¼
By means of depreciation the residual book values of assets are continued over the period. Together with acquisition and production costs and calculated equity ratio the residual book values are used to calculate the imputed return on equity. In cost category “Other Expenses”, only the wastewater discharge fees are quantified by the model. The wastewater discharge fees are calculated with the currently valid fee per pollution units of 35.79 V (AbwAG x 9). Pollution units are calculated in module WWTP. The remaining expenses of this cost category considered together and are depending on p.e. The sum of the cost categories yields in chargeable costs. From chargeable costs the reversal of subsidies and contributions are deducted and results in costs set equal to expenses from which the charges are calculated.
3. Results and discussion of the model validation and usage by an exemplary utility To verify the accuracy of the calculation algorithms, the model was tested by a validation using historical data from an exemplary utility. The utility is situated in the federal state MecklenburgVorpommern in the North-East of Germany. Mecklenburg Vorpommern has about 1.6 million inhabitants and is characterized by rugged settlement structures within large rural areas with a population density of 69 inhabitants per square kilometer and a settlement density of 849 inhabitants per square kilometers. This state is the most sparsely populated state in Germany and is characterized by demographic challenges. The population decreased by 16% from 1990 to 2015. In the future, it is expected that the population decreases further by 10% due to aging (StatA MV, 2013). After the political changes, the sewage infrastructure in this area was modernized and centralized to a great extent in the beginning of the 90's. The proportion of households that were connected to public sewage increased from 65% in 1990 to 88% in 2013 (LUNG, 2015). The entire state is equipped with a rather new sewer network because of the previous constructions and renovations. A special feature of the exemplary wastewater utility is the unusually high occurrence of business and industry compared to the rest of the region. Total served area of the utility 60,915 hectares (of which 18,218 hectares is urban area consisting of villages and three small towns). The assets comprise 221 km sewer networks and 14 WWTP with a total capacity of 206,500 p.e. (of which are 19,600 centrally connected residents, while the remaining load is due to industrial inflow). Because the utility was founded in 2001, the longest possible period to acquire technical and financial data for validation was between 2001 and 2014. But even within that period, the collection of technical data was not complete. So, the results of the module sewer cannot be compared with the actual development of sewer
Sewer cleaning Sewer renovation Sewer network Sewer network
systems state because of missing historical data of the sewage networks. As estimate, the statistic distribution of MV was applied (MLUV, 2009). With consultation of the exemplary wastewater utility, these values are to be regarded as representative. For this parameters only economic data can be compared, keeping the assumption based uncertainty in mind. Also the data from WWTPs were not complete, but were adequate enough to compare the results of the module WWTP. Furthermore, during the regarded period, resolution of subsides and contributions started in 2006. Changes in the financing structure during the simulation cannot considered by the model. For that reason, the charge resulting from the simulation must be subsequently corrected by the charge reducing reversals of connection charges and subsides. Technical and financial data from 2001 served as the starting values for model validation. External changing conditions and utility-intern key figures are related accordingly to this year. To compare model based prognosis with the actual development of the utility, the year 2014 was chosen, which provided the best data availability. Based on data from the utility and Federal Office of Statistics the following average annual assumptions for the simulation have been made:
Condition rates: Industrial growth rate: Population rate: Sewer development rate: Sewer rehabilitation rate: Sewer cleaning rate:
±0.0% 1.4% þ0.7% þ0.7% þ5.0%
Economic rates: Consumer cost rate: Construction costs: Electricity price: Grid feed-in tariff: Personnel cost rate: Sewer cleaning costs: Sewer rehabilitation costs: Sludge disposal costs:
þ1.7% þ1.6% þ10.5% 1.2% þ2.3% þ2.5% þ0.7% þ5.0%
3.1. Validation of process based parameters In 2014, consumed precipitants, produced sludge and energy consumption were given for all WWTP. The calculated demand of precipitants at the end of this 13 year long “prognosis period” is slightly underestimated by 4,7% (i.e. 1.527 kg Fe). The comparison of calculated and real values for dry sludge mass and electricity consumptions in the all WWTP operated by the utility is summarized in Fig. 2. In sum for the whole utility the simulation deviates only by þ2.6% from the actual values The largest from 14 WWTP of the utility, which has a design capacity of 200,000 p.e. and is as only one equipped with an anaerobic stabilization makes 95% of total sludge mass. Calculated value of this plant differs from the measured data by 3.1% (1226 m3 a1). Deviations in the small rural
56
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
Fig. 2. Comparison between measured and simulated sludge (left) and electricity consumption (right) of the different WWTP types: PTP-Pond Treatment Plant, RBC-Rotational Biological Contactor, BTP-conventional Biological Treatment Plant, BTPanaerob-conventional Biological Treatment Plant with anaerobic sludge stabilization.
Fig. 3. Electricity balance between calculated and measured values of electricity consumption and production.
plants are higher: for activated sludge systems 13% (43 m3 a1) and for the near natural pond systems by 11% (25 m3 a1). All the electric energy demand is met well. In sum of all WWTP the demand deviates by 2.9%. Non aerated ponds (PTP) consume no electricity (only for pumping) and are not shown. The large WWTP consumes 96% of the total electricity consumption of all WWTP. Calculated values of this plant deviate by 2.0% (41,513 kWh a1). The average deviation of the remaining WWTP (all activated sludge systems with simultaneous aerobic stabilization) is more significant 34% (5008 kWh a1), which may be due to case specific operational reasons (like aging of membranes, aeration control) or changes of the connected inhabitants on a smaller scale than the model resolution. The calculated and measured electric energy balance of the large WWTP with anaerobic stabilization is shown in Fig. 3. The measured annual electric energy demand of the large plant with anaerobic stabilization is 2,091,972 kWh. Thereof, the WWTP imported only 14,443 kWh, the remaining demand is covered by
internal power supply. In total, 4,397,982 kWh of electric energy are produced. The resulting surplus is 2,306,010 kWh and is fed into the grid. In the model, external energy supply for this plant is not required, because it assumes that the generated energy by sludge digestion is primarily used for self-consumption. The calculated electricity generation is with 4,244,724 kWh slightly lower than measured (3.5%). Since the model subtracts the whole own demand the grid feed deviates more 2,194,265 kWh. The occasional not available power supply by the own generator, e.g. due to maintenance could be estimated from operators experience and improve the model in this issue. 3.2. Validation of cost accounting The calculated expenses, derived from the calculated technical data, defined accounting conditions and economic rates compared to real expenses 2014 in Table 3. Generally, the model predicts well the development of the
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra Table 3 Comparison between detected and calculated costs from cost categories operating and purchased services. Cost/income
Electricity costs Electricity income Precipitant costs Sludge disposal costs Sewer cleaning costs Sewer repair costs
Table 4 Comparison of real and calculated cost categories. Cost category
D Calc to Real (V)
(%)
þ3001 þ2728 337 þ1517 1081 422
þ1.8 þ1.3 0.9 þ0.9 9.0 5.1
D Calc to real: Difference between calculated and real costs.
economic data. Despite calculated lower power consumption and energy generation (which is slightly compensated with respect to measured values) electricity income and electricity costs has a deviation of þ1.8% and þ1.3%. Regarding to higher simulated electricity costs it's possible that it depends on individual contracts for energy supply, which are lower than market-standard. As expectable from the good agreement for the technical data, expenses for precipitants and sludge disposal deviated with 0.9% and þ0.9%, respectively only slightly from the real expenses. The deviations of sewer cleaning costs and sewer repair costs are 9.0% and 5.1%. Underestimations of both costs can have different reasons. Either the specific costs or the calculated sewer lengths for cleaning or repair, or both can be too less. With regard to the idealized approaches for aging, rehabilitation and cleaning demand, the estimations of CS of the sewer system and the not calibrated cohort survival functions, the accuracy is satisfying. The annual cost structure (Fig. 4) shows the change in the respective cost categories. In the simulation total costs increased by þ9.2%. For a better comparison, percentages of real and calculated changes and percentage of deviation between real and calculated costs are shown in Table 4. Personnel costs were simulated precisely, the difference between real and calculated values is 360 V (0.03%). By taking into account of contrary price developments at electricity price and precipitation price the development of operating costs are well described, the deviation is 16,881 V (3.84%). Despite minor deviations at sludge disposal costs, sewer cleaning costs and sewer repair costs the largest deviation is at purchased services with 48,698 V (3.79%). The reason may be the dependency of
57
D Calc to real (V)
(%)
Personnel Operating Purchased services Interest Other expenses Depreciation
360 16,881 48,698 þ39,637 þ2071 þ1446
0.03 3.84 3.79 þ5.84 þ0.48 þ0.08
Total costs
22,785
0.35
D Calc to real: Difference between calculated and real costs.
Table 5 Differences between real and calculated amount charge under considering of the accounted reversals. Item
Amount Charge (V m3)
Calculated volume charge Reversal of subsides and contributions
þ2.94 0.35
Adjusted volume charge
þ2.59
rest expenses of cost category on p.e. Expenses do not actually decrease in the same dependency on p.e. as in the model. The largest percentage deviation is at interest with þ5.84% (þ39,637 V). A reason for the deviation can be too high interest rates. If applicable, the utility received loans at better conditions. Calculated cost categories other expenses and depreciation has a high accuracy with a deviation of þ2071 V (þ0.48%) and þ1446 V (þ0.08%). Due to the fact that deviations at depreciation are low, it can be assumed that specific prices for sewer rehabilitation and renewal or the cost functions for the reinvestment of the WWTP are sufficiently precise. Resulting on larger deviations revise each other a minor deviation at total costs of 0.35% (22,785 V) occurs. Usually, rising expenses must lead to increasing basic or volume charge. For a simplified consideration and a better comparison of the results the basic charge is set to constant. In this way, changes in costs will be allocated to the volume charge. The resulting cost structure from the simulation and changing volume charges are depicted in Fig. 4. Within the simulated period, 2001 to 2014 a required increase by þ0.34 V from 2.60 V to 2.94 V (13.3%) was calculated. But, these results still have to be adjusted with respect
Fig. 4. Simulated annual cost structure (left) and simulated annual volume charge (right).
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
58
to accounted reversal of subsidies and contribution (see Table 5). Compared to the real volume charge of 2014 the prognosis just deviates by 1 cent m3. Multiplied with the wastewater volume, the resulting proceeds via the charge are underestimated by 22,785 V. Considering the total utilities revenues statement of 2014, the value corresponds to an underestimation of 0.4%. 3.3. Application for forecasting the development till 2050 After successful validation, the model is exemplarily applied to forecast the development of the utility starting from 2014 until to 2050. For this, technical and economic data as well as the given basic and volume charge are used. Conditions for this period are defined by the wastewater utility. The following assumptions were made:
Condition rates: Industrial growth rate: Population rate: Sewer development rate: Sewer rehabilitation rate: Sewer cleaning rate:
±0.0% 0.5% ±0.0% þ2.0% þ5.3%
Economic rates: Consumer cost rate: Construction costs: Electricity price: Grid feed-in tariff: Personnel cost rate: Sewer cleaning costs: Sewer rehabilitation costs: Sludge disposal costs:
þ2.0% þ2.0% þ2.0% ±0.0% þ1.5% þ2.0% þ2.0% þ2.0%
On the basis of the defined conditions, the resulting development of expenses at each cost category is illustrated in Fig. 5. With regard to the development of personnel costs of þ1.5% per year and the same number of employees, personnel costs would rise by þ793,862 V (60.0%) until the end of the period. As a result of decreasing population, electricity consumption of sewage treatment plants would reduce by 18.2%, electricity generation by 2.4% and precipitation consumption by 11.6%. However, in spite of decreasing operating materials, due to inflation expenses of the cost category would increase by þ67,618 V (þ16.0%). The cost of purchased services would increase by þ517,571 V (38.1%), including þ198,604 V (51.6%) for disposal of sewage sludge, þ62,026 V (þ775.3%) for sewer repairs and further þ5973 V (þ79.6%) for sewer cleaning. Regarding sewage sludge, the mass would decrease by 11.9%, but caused by inflation and the consideration that in future sludge income will (for legal reasons) probably be incinerated instead of agricultural valorization the costs for sewage sludge disposing would increase. The expenses of
Fig. 6. Simulated annual volume charge.
sewer repairs would increase despite an annually renovation rate of 2%. This is due to the special homogeneous age structure of the sewer system, which leads despite the high investments in rehabilitation and renovation to deteriorating sewer conditions. Furthermore, expenses for interest would rise by þ236,977 V (þ32.3%), other expenses decreased by 47,442 V (9.1%) and depreciation would rise by þ98,529 V (þ5.4%). The charge resulting from the sum of expenses, reduced by resolutions, is shown in Fig. 6. On the basis of the given conditions the volume charge would increase by þ1.23 V (þ47.4%) from 2.60 V to 3.83 V. By the reversal from the residual value of contributions in 2021 the calculated charge jumps. The charge reduction in 2023 as well as a stagnation in 2045 result from the reinvestments of WWTP, which are dimensioned to reduced p.e. at time of investment. 4. Conclusion The developed simplified process-based model for technical and economic quantification can close a gap between desired technical and economic forecasting based on process based relationships and the necessity to simplify those forecasts for reasons of applicability
Fig. 5. Simulated annual cost structure from 2014 to 2050.
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
and data availability. The model has been developed in close cooperation with several utilities and addresses their requirements. The validation with historical data provided by an exemplary utility (time series of 13 years) has shown that the results reflect well both, technical key data and derived current costs and charges. While the calculation approaches for comparable technical systems are transferrable (which is the case for many European regions), the cost accounting model may need adaptations when transferring it to different accounting rules. The accuracy of the model can be improved by case specific definition of calculation parameters. This applies namely for probably cost sensitive calculations like the cohort survival functions in module Sewer. In the case given, parameter for this function literature value had to be used (Jansen, 2007), due to missing data. The process oriented setup of the model enables its user to assess widely varying development scenarios and technical as well as organizational management options showing differentiated causal relationships. Acknowledgements The study was funded by a state graduate scholarship of the state Mecklenburg Vorpommern, Germany. We would like to thank all stakeholders and ministries for their cooperation and support during the investigations. References €sser e AbwAG, 2005. Gesetz über Abgaben für das Einleiten von Abwasser in Gewa Abwasserabgabengesetz (Law on charges for the discharge of wastewater into waters e Wastewater Levy Act). Germany, Version of the Announcement from 18.01.2005. Ana, E.V., Bauwens, W., 2010. Modeling the structural deterioration of urban drainage pipes: the state-of-the-art in statistical methods. Urban Water J. 17, 47e59. ATV - DVWK - A 131, 2000. Bemessung von einstufigen Belebungsanlagen (Dimensioning of single-stage activated sludge plants). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. €rpern und RotationATV - DVWK - A 281, 2001. Bemessung von Tropfko € rpern (Dimensioning of trickling filters and rotating biological constauchko tactor). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. BDEW, 2015. Industriestrompreise: Energieinfo - Ausnahmeregelungen bei Energiepreis-bestandteilen (Industrial electricity prices: Energy information - Derogations for energy price components). Bundesverband der Energie- und Wasserwirtschaft e. V., Berlin, Germany. BMWi, 2016. EEG in Zahlen: Vergütungen, Differenzkosten und EEG-Umlage 2000 bis 2017 (EEG in numbers: Remuneration, differential costs and EEG allocation 2000 to 2017). Bundesministerium für Wirtschaft und Energie, Berlin, Germany. Castellet, L., Molinios-Senante, M., 2016. Efficiency assessment of wastewater treatment plants: a data envelopment analysis approach integrating technical, economic, and environmental issues. J. Environ. Manage. 167, 160e166. €tze für Bemessung, Bau und Betrieb von AbwasserteiDWA-A 201, 2005. Grundsa chanlagen (Principles for dimensioning, construction and operation of wastewater treatment ponds). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. DWA-M 149-3, 2015. Zustandserfassung und -beurteilung von Entw€ asser€uden e Teil 3: Beurteilung nach optischer ungssystemen außerhalb von Geba Inspektion (Conditions and Assessment of Drain and Sewer Systems Outside Buildings - Assessment by visual inspection). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. €sserung (Mechanical sludge dewDWA-M 366, 2000. Maschinelle Schlammentwa atering). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. Ferreira da Cruz, N., Carvalho, P., Marques, R.C., 2013. Disentangling the cost
59
efficiency of jointly provided water and wastewater services. Util Policy 24, 70e77. Gretzschel, O., Schmitt, T.G., Hansen, J., Siekmann, K., Jakob, J., 2014. Sludge digestion instead of aerobic stabilisation - a cost benefit analysis based on experiences in Germany. Water Sci. Technol. 69, 430e437. Haider, H., Sadiq, R., Tesfamariam, S., 2016. Intra-utility performance management model (In-UPM) for the sustainability of small to medium sized water utilities: conceptualization to development. J Clean Prod. 133, 777e794. Henze, M., Gujer, W., Mino, T., van Loosdrecht, M.C.M., 2000. Activated Sludge Models ASM1, ASM2, ASM2d and ASM3. IWA Task Group on Mathematical Modeling for Design and Operation of Biological Wastewater Treatment. IWA Publishing, London, England. Hernandez-Sancho, F., Molinos-Senante, M., Sala-Garrido, R., 2011. Cost modeling for wastewater treatment processes. Desalination 268, 1e5. Herz, R., 1996. Aging processes and rehabilitation needs of drinking water distribution networks. J. Water SRT Aqua 45, 221e231. Imhoff, K., Imhoff, K.R., 2007. Taschenbuch der Stadtentw€ asserung (Paperback of urban drainage). Oldenbourg-Verlag, Munich, Germany. Jansen, K., 2007. AQUA-WertMin Program Manual. Expert Office for Sewer, Klienbittersdorf, Germany. KAG M-V, 2005. Kommunalabgabengesetz e KAG M e V (State Municipal Charges Act of the state Mecklenburg e Vorpommern). State MecklenburgVorpommern, Germany, Version of the Announcement from 12.04.2005. Kleiner, Y., Rajani, B., 2001. Comprehensive review of structural deterioration of water mains: statistical models. Urban Water 3, 131e150. Kley, G., Caradot, N., 2013. Review of Sewer Deterioration Models e Project Acronym: SEMA. Report from Kompetenzzentrum Wasser Berlin gGmbH. Department Surface Water, Berlin, Germany. Le Gat, Y., 2008. Modelling the deterioration process of drainage pipelines. Urban Water J. 5, 97e106. LUNG, 2015. Kommunale Abwasserbeseitigung in Mecklenburg-Vorpommern: €ß Richtlinie 91/271/EWG (Sewage disposal in Lagebericht 2015-Bericht gema Mecklenburg-Vorpommern: Status report 2015 e Report according to Directive 91/271/EEC). Landesamt für Umwelt, Naturschutz und Geologie MecklenburgVorpommern, Güstrow, Germany. MLUV, 2009. Zustand der Abwasserkanalisation in Mecklenburg-Vorpommern (Conditon state of sewer networks in Mecklenburg-Vorpommern). Ministerium für Landwirtschaft, Umwelt und Verbraucherschutz Mecklenburg-Vorpommern, Schwerin, Germany. Rajani, B., Kleiner, Y., 2001. Comprehensive review of structural deterioration of water mains: physically based models. Urban Water 3, 151e164. Rehan, R., Knight, M.A., Unger, A.J.A., Haas, C.T., 2014. Financially sustainable management strategies for urban wastewater collection infrastructure e development of a system dynamics model. Tunn. Undergr. Sp. Technol. 39, 116e129. Ruiz-Rosa, I., García-Rodríguez, F.J., Mendoza-Jimenez, J., 2016. Development and application of a cost management model for wastewater treatment and reuse processes. J. Clean Prod. 113, 299e310. Silva, C., Saldanha Matos, J., Rosa, M.J., 2016. Performance indicators and indices of sludge management in urban wastewater treatment plants. J. Environ. Manage. 184, 307e317. €lkerStatA MV, 2013. Aktualisierte 4. Landesprognose - Basisjahr 2010: Bevo ungsentwicklung des Landes Mecklenburg e Vorpommern sowie 48 der €dte und Landkreise bis 2030 nach Altersgruppen (Updated 4th kreisfreien Sta country forecast - base year 2010: Population change of Mecklenburg e Vorpommern as well as 48 of the independent cities and counties until 2030 by age groups). Statistical reports. Bd. A 1831 2012 01, State Statistical Office of Mecklenburg e Vorpommern, Schwerin, Germany. StatBA, 2016. Preisindizes für die Bauwirtschaft (Price indices for the construction industry). Statitisches Bundesamt, Fachserie 17 Reihe 4, Artikelnummer: 2170400163244, Wiesbaden, Germany. StatBA, 2017. Preise: Verbraucherpreisindizes für Deutschland - Lange Reihen ab 1948 (Prices: Consumer price indices for Germany - Long series from 1948). Statitisches Bundesamt, Artikelnummer: 5611103161124, Wiesbaden, Germany. €nckner, J., Franz, T., Frehmann, T., Jathe, R., Obermayer, A., Winkler, U., 2014. Tra Wirtschaftliche Auswirkungen sich überlagernder Entwicklungstrends auf Abwasserentsorgungsunternehmen (Economic effects of overlapping trends on wastewater utilities). KA Korrespondenz Abwasser 61, 793e801. Yang, J., 2004. Road Crack Condition Performance Modeling Using Recurrent Markov Chains and Artificial Neural Networks. Doctoral Dissertation. Department of Civil and Environmental Engineering, College of Engineering,University of South Florida.
Contents lists available at ScienceDirect
Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman
Research article
A combined model to assess technical and economic consequences of changing conditions and management options for wastewater utilities €nckner b Mathias Giessler a, b, *, Jens Tra a b
Emschergenossenschaft/Lippeverband, 45128 Essen, Germany University of Rostock, Department for Water Management, 18059 Rostock, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 April 2017 Received in revised form 22 October 2017 Accepted 6 November 2017 Available online 12 December 2017
The paper presents a simplified model that quantifies economic and technical consequences of changing conditions in wastewater systems on utility level. It has been developed based on data from stakeholders and ministries, collected by a survey that determined resulting effects and adapted measures. The model comprises all substantial cost relevant assets and activities of a typical German wastewater utility. It consists of three modules: i) Sewer for describing the state development of sewer systems, ii) WWTP for process parameter consideration of waste water treatment plants (WWTP) and iii) Cost Accounting for calculation of expenses in the cost categories and resulting charges. Validity and accuracy of this model was verified by using historical data from an exemplary wastewater utility. Calculated process as well as economic parameters shows a high accuracy compared to measured parameters and given expenses. Thus, the model is proposed to support strategic, process oriented decision making on utility level. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Process-based model Economic quantification Wastewater utilities Changing conditions
1. Introduction In contrast to the global trend many regions in Europe suffer from a decrease and aging of population, often described as “demographic change”. In parallel, legal, technical and economic conditions are changing with integrally acting consequences for utilities. Assessment of the effects and development of appropriate adaptation measures demand for approaches which are able to describe the causal and complex relationships between changing conditions, technical system and economic effects. The shrinking population, high fix costs and long depreciation periods create economic but also technical challenges. To adapt timely and sustainable, it is important to quantitatively predict technical and economic effects of those changes and to evaluate the effectiveness of subsequent measures. There exist a lot of valuable programs for simulating processes of WWTP, sewer system as well the aquatic environment with which the effects of changing conditions can be considered. These programs are generally focused on certain systems and questions (like hydrodynamic sewer models, rehabilitation models, activated sludge models etc.). Generally, they are complex and require a high data density and thorough
* Corresponding author. Emschergenossenschaft/Lippeverband, 45128 Essen, Germany. E-mail address: [email protected] (M. Giessler). https://doi.org/10.1016/j.jenvman.2017.11.016 0301-4797/© 2017 Elsevier Ltd. All rights reserved.
parameterization. Even cost management models for efficiency analysis and decision support tools provide no holistic approach to reflect the cost of utility administration, sewer and treatment plants. Different approaches for both sectors are presented in literature. For sewer maintenance cost, a dynamic systems model developed by Rehan et al. (2014) to support the development of financially sustainable management strategies. Hernandez-Sancho et al. (2011) developed a cost modeling methodology for wastewater treatment processes based on statistical information by aiming at a better understanding of the cost structure of wastewater treatment processes. Also based on data envelopment analysis (DEA) Castellet and Molinios-Senante (2016) developed a nonradial approach by integrating technical, economic and environmental issues for efficiency evaluation of WWTPs. Ruiz-Rosa et al. (2016) designed and adapted a cost management model to calculate the costs of each activity involved in wastewater treatment and reuse processes by combining DEA and Life Cycle Assessment (LCA) indicators. On utility level, models exist to estimate the performance and efficiency. One example is the DEA model by Ferreira da Cruz et al. (2013) which measures separately the efficiency of each service. In combination with other indicators, this model can be used for prioritizing efforts to improve overall efficiency. An intrautility performance management model was proposed by Haider et al. (2016) as a decision tool for sustainability assessment of small and medium sized water utilities.
52
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
An integrated technical and economic planning, especially if changing boundary conditions are regarded, a process-oriented approach is required. This has to combine models, which realistically can describe the causal relationships within the actual technical system with an economic assessment tool, reflecting the cost structure of the respective utility. A rather pragmatic approach for such a combined assessment has been proposed by the DWA-Workgroup WI-1.3 and illustrated for a defined “standard water utility”. This method based on a simplified simulation that uses corporate benchmarking indicators. For selected trends, the economic effects of and their superposition can be derived. The model serves to generalize economic effects on changing conditions in the water sector, but depends on the availability of utility specific benchmarking data. The relationships between changing boundary conditions, infrastructure, and economic effects are rather conceptual. Complex interactions or e.g. changes in organizational structures and far going changes of the €nckner et al., technical system cannot be adequately reflected (Tra 2014). To achieve a balance between limited data availability and expressiveness a process oriented but still sufficiently simple model was developed. This paper presents The structure of the simplified process-based and economic model focused on publicly available infrastructure data, WWTPs data and expenses of the utilities. The validation of the model based on historical data of an exemplary utility in Mecklenburg-Vorpommern, Germany. An exemplary quantitative forecast of technical and economic effects till 2050 due to assumed future changes. The following sections describe the model development and its structure as well as its components (modules) and input parameters. Based on data from an exemplary utility, the model is tested, validated and the results and accuracy are discussed. 2. Model development and structure The model is programmed in MATLAB®, a commercial numerical computing environment provided by MathWorks®. It comprises substantial cost relevant assets and activities of a typical German wastewater utility limited to the core tasks wastewater disposal and treatment. It is subdivided into three modules (see Fig. 1). The module Sewer is a simplified asset management tool of the sewer system, modeling the condition as function of aging and defined activities for repair/renovation/renewal. The module WWTP is a process oriented tool calculating inflow and outflow parameters of the actual specific WWTP as well as cost relevant technical parameters. All necessary parameters from the module Sewer and WWTP are used in economic assessment module Cost Accounting. 2.1. Data set In order to specifically address the different structural, operational and administrative conditions of a specific utility, the model demands a rather comprehensive data set. In the phase of model development, data collection was rather challenging due to the data's sensitivity and the required time-consuming preparation. Therefore, previously collected data provided by statistical institutes and ministries were accessed. Table 1 gives an overview of the collected data and their origins. The data of statistical institutes and ministries were taken at municipality level. The service area of a utility often differs from the official administrative borders. This required structured and reproducible approaches to compare process and assign the actual
service area to available data. For the case study, the data and their interrelations were double-checked for accuracy and plausibility by the stakeholder. In addition, individual surveys were conducted to obtain stakeholder data.
2.2. Module sewer The module Sewer describes the development of sewage systems state due to aging and subsequent rehabilitations measures. Input parameters for the module sewer are i) the classification of the sewer network into age groups (<10a, >10a, >20a, >30a, >40a, >50a) and ii) condition states (CS) according to DWA-M149-3 (2015) from CS 5 (best condition, meaning new sewers with no defects) to CS 0 (worst condition, meaning old sewers at the end of their life cycle with many or great defects). Calculation of age groups, determines the end of depreciation from network parts. Due to structural deterioration, the sewage system does not develop linearly, rehabilitation may be necessary at earlier stages. Therefore, CS are estimated to determine which parts of the sewer network need rehabilitation. The sewer system is rehabilitated according to age and condition, beginning with the oldest age group and worst CS and ending with the younger age groups and better CS. The rehabilitations and developments are calculated in annual time steps for the cost calculation while transitions between age groups occur in ten years intervals. For transition of CS, the cohort survival function from Herz (1996) is used. This function, called the Herz-distribution, calculates the annual percentage of sewer parts which stay in the same CS (cohort). The parameters of the survival functions, including the aging vector, the transition vector and the resistance vector, were calibrated according to Jansen (2007), who describes the calibration for the program AQUA-WertMin. The parameters can be adjusted to the specific sewer system. This approach is also integrated in other software products like DynaStrat and KANEW-Z, which are mainly proposed by German consulting offices (Kley and Caradot, 2013). In context of long term strategies and budget requirements, the cohort survival approach is very useful and is simple to compute (Ana and Bauwens, 2010; Kley and Caradot, 2013). However, the Herz-functions seem to overestimate the actual survival rates and remaining life expectancies (Le Gat, 2008). Beside the statistical deterioration approach from Herz, there are countless different deterioration models which could alternatively implemented. Several review papers give a good overview of the available models. Yang (2004) categorized the model types into three classes: physical models, artificial intelligence based models (e.g neural network, Fuzzy set theory), and statistical models. Rajani and Kleiner (2001) give an overview and description of physical models. They subdivided them into probabilistic and deterministic models. The companion paper of Kleiner and Rajani (2001) focuses on statistical models and divide this class into deterministic models with a subdivision in time exponential models and time linear models as well as probabilistic models with the subdivision in multi-variate models (proportional hazards models, accelerated lifetime models, time-dependent poisson models) and singlevariate group-processing models (cohort survival model, Bayesian diagnostic model, semi-markov model, break clustering). Ana and Bauwens (2010) subdivide the statistical models generally into two groups: pipe group models to predict the condition of a sewer group (cohorts) and pipe level models to simulate each single pipe. For all groups different models were presented and described in the paper. Because of the data set in this study, only statistical sewer group models can be used. Models to simulate individual pipes need a more detailed data set.
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
53
Fig. 1. Modular schema of the created model.
Table 1 Collected data with their origin. Data
Source
Type of survey
Dataset
Sewer length Sewer age Kind of sewage disposal (centralized/ decentralized) Population equivalent Sewage (type, amount) Population Areas according to land utilization Number of buildings and apartments
Statistical State Office MV (StatA MV)
Federal Statistics
structural sewer data
Design capacity Process Technology Construction year and capacity Nominal load In- and Outflow parameters (Q,COD,N,P) Permit values (COD,N,P)
State Office for Environment, Nature Conservation and Geology MV (LUNG)
Survey of the LUNG base on Wastewater Levy Act (x 13 (1) AbwAG)
P Elimination Sludge treatment/disposal CoSubstrate Energy consumption/production Charge Annual report (balance sheet, income statement) Personnell structure Levying/Reversal of contribution Funding/release subsidies Asset values
Wastewater utility
Own survey
structural data municipality
Micro census
2.3. Module WWTP Module WWTP serves to simulate effects of changing conditions in wastewater treatment plants. Running a dynamic process model as the activated sludge models (ASM) 1 to 3 (Henze et al., 2000) is from the perspectives of effort for setting up the model, the data availability and desired time resolution not advisable. Instead a numerical mass-flow model, based on annual input data was developed. Based on quantified mass flows, operational costs calculated.
Object data WWTP
Economy data
Module WWTP simulates sludge production (primary sludge, activated sludge), demand of precipitants and flocculants, sludge treatment and disposal, energy production of digester gas and electricity consumption of the plant. Therefore, conventional WWTP- including nutrient removal and the option of anaerobic sludge stabilization as well as WWTP with aerobic sludge stabilization and the option of anaerobic pretreatment for special inflow conditions are taken into account. Based on ATV-DVWK A 131 (2000) Annex 2, the activated sludge system is calculated basically by combining of chemical oxygen demand (COD), nitrogen (N)
54
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
and phosphorous (P) mass balances with kinetic functions of growth and decay. So far, average standard values of the guideline are used for calculation but can be replaced if more detailed information is available. Sludge production in rotational biological contactor (RBC) plants and trickling filters is calculated according to ATV - DVWK - A 281 (2001). For primary settlers with one hour detention time, ATV-DVWK A 131 (2000) proposes an efficiency COD 35%, total suspend solids (TSS) 60%, N 10% and P 10%. For pond treatment systems, the cleaning performance for organic matter of a tailing pond as a primary clarifier stage is according to DWA-A 201 (2005) assumed with 30e50%. WTTP's with a design capacity 10,000 population equivalent (p.e.) are in Germany often designed for simultaneous aerobic sludge stabilization without primary clarification and as default described in the model accordingly. Gas yield and, subsequently, the generated energy are calculated according to Gretzschel et al. (2014). This calculation depends on the sludge conditions and the presence of co-substrates. The reduction of sludge by anaerobic treatment is taken into account with a factor by Imhoff and Imhoff (2007). Estimation of sludge mass to be disposed depends on dewaterability. Values for this parameter in conjunction with sludge as well as process technology (aerobic or anaerobic) and procedural steps (conditioning, dewatering, drying) like in the presented state-variables selected for their relevance to WWTP sludge management according to Silva et al. (2016), were taken from DWA-M-366 (2000). In comparison to that model, module WWTP has no evaluation of the sludge management performance. Energy consumption of treatment processes are determined by p. e. specific consumption depending on process technology. For plants which produce electricity through sludge digestion, the model assumes that electricity production covers the plants own energy demand first, leading an according reduced net consumption. If due to special conditions more energy is produced than required the surplus is fed into the grid. The calculation of pollution units for wastewater discharge fees is based on German wastewater Levy Act x3 (AbwAG, 2005). COD, N and P are regarded, while heavy metals are neglected due to missing data. 2.4. Module cost accounting The module Cost Accounting is based on the State Municipal Charges Act (KAG M-V, 2005) and results in the cost structure of the utility as well as derived charges to the wastewater producers. Beside the above described consideration of the actual technical system and technology, it takes the utility specific organizational and financial structures into account. Results from the modules WWTP and Sewer converge in this calculation, and are accounted and allocated to the respective cost category. All expenses are summarized and represented by the cost structure. Charge-reducing liquidation of contributions and subsidies are, if given, counted and updated. If sewers are newly connected, an actual contribution rate is determined in accordance with the KAG M-V x 7. Considering the investments of WWTP and sewers, resulting interest through debt financing and depreciation, future assets and liabilities are simulated. All information ends in the calculation of charges under condition of cost coverage. The rules for cost accounting and charging are differently set, even by the federal states within Germany. Application of the model in other regions requires according adaptions. Table 2 shows how chargeable costs are composed and which costs of the cost category have been considered by technical parameters in the respective cost centers. All costs in the model are subject to inflationary influences.
Unless otherwise indicated, the rise of consumer prices was taken as inflation. With regard to contrasting or intensified trends in price developments compared to the consumer prices (StatBA, 2017) in the past, separated inflation rates for electricity (BDEW, 2015), grid feed (BMWi, 2016), sludge disposal (requested by the utilities) and sewer cleaning (requested by the utilities) are used. For the construction of sewage systems (sewer, WWTP) the construction index for civil engineering (StatBA, 2016) has been set. Generally the personnel cost in the model results on the number of employees and wage developments. This cost category is divided into administration, WWTP and sewer. Due to different structures in the organization, specific ratios for the detection of the required number of employees in the administration are too inaccurate. Furthermore, certain processes, such as sewer rehabilitation and cleaning or sludge disposal can also be outsourced. In the model, these costs are in the cost category purchased services, so they not directly influence the personnel costs. Prospected personnel strength at the end of the simulation period is estimated and must be indicated. For WWTP-staff, the workload for maintenance and operation is taken into account according to the state ordinance on self-monitoring of sewage plants and sewage discharges. In case of a change in the WWTP size class or process technology, the increase or decrease in expenditure by man-hours is counted to personnel costs. Electricity costs are calculated by electricity consumption, electricity price considering the cost increase. The costs for precipitants are based on consumed mass, price and inflation rate for consumer prices. Incomes from surplus of electrical energy which are fed into grid are invoiced with investments. The remaining share of operating costs has been summarized and due to this share is mainly composed of costs from sewer network operation it was assumed that these costs are changing proportionally to dry weather flow. Sludge disposal costs are accounted as purchased services. The costs depend strongly on the disposal way (agricultural or incineration). Prices are defined per cubic meter to be disposed and a derived specific price increasing rate. Costs for sewer network repairs results from the CS of the sewer network in module Sewer. Only CS 0 to 2 are declared as need for repair. For each of these three CS, a specific price is determined per meter. The worse sewer condition the higher is expense of repair cost. The remaining expenses in cost category purchased services are summarized and proportionally to connected p. e. The calculation of the cost categories depreciation and interest is constructed in a functionally same manner. The total amount of depreciation and interest results from the shares of WWTP and sewer as well as the summary of other tangible assets under others. Depreciation is distinguished between imputed depreciation and depreciation of acquisition and production costs. The kind of depreciation can be selected by the user. According to KAG M-V x 6 (2a), linear depreciation is applied. In addition to the acquisition and production costs, assets to be reinvested are also included in the calculation. The asset age is taken from the modules WWTP and Sewer and the end-of-service life is determined. In the case of reinvestment (depending on the strategy), the replacement value is calculated at the respective point in time. From this investment value, interest and depreciation are calculated and added to the respective cost types. The remaining asset-values, which are not part of WWTP or sewer network and not given in the data, are summarized in one position. For the model, it was assumed that this position mainly contains administrative assets (buildings, vehicles and others). By reason of the dominant share of buildings the asset-value, the depreciation period of this position was set equal to that of buildings with 50 years.
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
55
Table 2 Considered parameters of cost categories within the cost centers. Cost category
Administration
WWTP
Sewer network
Personnel Operating Purchased services Depreciation Interest Other expenses
Personnel costs for administration
Personnel costs for WWTP Electricity Precipitant Sludge disposal WWTP WWTP Wastewater levy
Personnel costs for sewer network
Buildings and others Buildings and others
e
P Chargeable costs Reversal of connection charges
e
Reversal of subsides
¼
Costs set equal to expenses
¼
By means of depreciation the residual book values of assets are continued over the period. Together with acquisition and production costs and calculated equity ratio the residual book values are used to calculate the imputed return on equity. In cost category “Other Expenses”, only the wastewater discharge fees are quantified by the model. The wastewater discharge fees are calculated with the currently valid fee per pollution units of 35.79 V (AbwAG x 9). Pollution units are calculated in module WWTP. The remaining expenses of this cost category considered together and are depending on p.e. The sum of the cost categories yields in chargeable costs. From chargeable costs the reversal of subsidies and contributions are deducted and results in costs set equal to expenses from which the charges are calculated.
3. Results and discussion of the model validation and usage by an exemplary utility To verify the accuracy of the calculation algorithms, the model was tested by a validation using historical data from an exemplary utility. The utility is situated in the federal state MecklenburgVorpommern in the North-East of Germany. Mecklenburg Vorpommern has about 1.6 million inhabitants and is characterized by rugged settlement structures within large rural areas with a population density of 69 inhabitants per square kilometer and a settlement density of 849 inhabitants per square kilometers. This state is the most sparsely populated state in Germany and is characterized by demographic challenges. The population decreased by 16% from 1990 to 2015. In the future, it is expected that the population decreases further by 10% due to aging (StatA MV, 2013). After the political changes, the sewage infrastructure in this area was modernized and centralized to a great extent in the beginning of the 90's. The proportion of households that were connected to public sewage increased from 65% in 1990 to 88% in 2013 (LUNG, 2015). The entire state is equipped with a rather new sewer network because of the previous constructions and renovations. A special feature of the exemplary wastewater utility is the unusually high occurrence of business and industry compared to the rest of the region. Total served area of the utility 60,915 hectares (of which 18,218 hectares is urban area consisting of villages and three small towns). The assets comprise 221 km sewer networks and 14 WWTP with a total capacity of 206,500 p.e. (of which are 19,600 centrally connected residents, while the remaining load is due to industrial inflow). Because the utility was founded in 2001, the longest possible period to acquire technical and financial data for validation was between 2001 and 2014. But even within that period, the collection of technical data was not complete. So, the results of the module sewer cannot be compared with the actual development of sewer
Sewer cleaning Sewer renovation Sewer network Sewer network
systems state because of missing historical data of the sewage networks. As estimate, the statistic distribution of MV was applied (MLUV, 2009). With consultation of the exemplary wastewater utility, these values are to be regarded as representative. For this parameters only economic data can be compared, keeping the assumption based uncertainty in mind. Also the data from WWTPs were not complete, but were adequate enough to compare the results of the module WWTP. Furthermore, during the regarded period, resolution of subsides and contributions started in 2006. Changes in the financing structure during the simulation cannot considered by the model. For that reason, the charge resulting from the simulation must be subsequently corrected by the charge reducing reversals of connection charges and subsides. Technical and financial data from 2001 served as the starting values for model validation. External changing conditions and utility-intern key figures are related accordingly to this year. To compare model based prognosis with the actual development of the utility, the year 2014 was chosen, which provided the best data availability. Based on data from the utility and Federal Office of Statistics the following average annual assumptions for the simulation have been made:
Condition rates: Industrial growth rate: Population rate: Sewer development rate: Sewer rehabilitation rate: Sewer cleaning rate:
±0.0% 1.4% þ0.7% þ0.7% þ5.0%
Economic rates: Consumer cost rate: Construction costs: Electricity price: Grid feed-in tariff: Personnel cost rate: Sewer cleaning costs: Sewer rehabilitation costs: Sludge disposal costs:
þ1.7% þ1.6% þ10.5% 1.2% þ2.3% þ2.5% þ0.7% þ5.0%
3.1. Validation of process based parameters In 2014, consumed precipitants, produced sludge and energy consumption were given for all WWTP. The calculated demand of precipitants at the end of this 13 year long “prognosis period” is slightly underestimated by 4,7% (i.e. 1.527 kg Fe). The comparison of calculated and real values for dry sludge mass and electricity consumptions in the all WWTP operated by the utility is summarized in Fig. 2. In sum for the whole utility the simulation deviates only by þ2.6% from the actual values The largest from 14 WWTP of the utility, which has a design capacity of 200,000 p.e. and is as only one equipped with an anaerobic stabilization makes 95% of total sludge mass. Calculated value of this plant differs from the measured data by 3.1% (1226 m3 a1). Deviations in the small rural
56
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
Fig. 2. Comparison between measured and simulated sludge (left) and electricity consumption (right) of the different WWTP types: PTP-Pond Treatment Plant, RBC-Rotational Biological Contactor, BTP-conventional Biological Treatment Plant, BTPanaerob-conventional Biological Treatment Plant with anaerobic sludge stabilization.
Fig. 3. Electricity balance between calculated and measured values of electricity consumption and production.
plants are higher: for activated sludge systems 13% (43 m3 a1) and for the near natural pond systems by 11% (25 m3 a1). All the electric energy demand is met well. In sum of all WWTP the demand deviates by 2.9%. Non aerated ponds (PTP) consume no electricity (only for pumping) and are not shown. The large WWTP consumes 96% of the total electricity consumption of all WWTP. Calculated values of this plant deviate by 2.0% (41,513 kWh a1). The average deviation of the remaining WWTP (all activated sludge systems with simultaneous aerobic stabilization) is more significant 34% (5008 kWh a1), which may be due to case specific operational reasons (like aging of membranes, aeration control) or changes of the connected inhabitants on a smaller scale than the model resolution. The calculated and measured electric energy balance of the large WWTP with anaerobic stabilization is shown in Fig. 3. The measured annual electric energy demand of the large plant with anaerobic stabilization is 2,091,972 kWh. Thereof, the WWTP imported only 14,443 kWh, the remaining demand is covered by
internal power supply. In total, 4,397,982 kWh of electric energy are produced. The resulting surplus is 2,306,010 kWh and is fed into the grid. In the model, external energy supply for this plant is not required, because it assumes that the generated energy by sludge digestion is primarily used for self-consumption. The calculated electricity generation is with 4,244,724 kWh slightly lower than measured (3.5%). Since the model subtracts the whole own demand the grid feed deviates more 2,194,265 kWh. The occasional not available power supply by the own generator, e.g. due to maintenance could be estimated from operators experience and improve the model in this issue. 3.2. Validation of cost accounting The calculated expenses, derived from the calculated technical data, defined accounting conditions and economic rates compared to real expenses 2014 in Table 3. Generally, the model predicts well the development of the
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra Table 3 Comparison between detected and calculated costs from cost categories operating and purchased services. Cost/income
Electricity costs Electricity income Precipitant costs Sludge disposal costs Sewer cleaning costs Sewer repair costs
Table 4 Comparison of real and calculated cost categories. Cost category
D Calc to Real (V)
(%)
þ3001 þ2728 337 þ1517 1081 422
þ1.8 þ1.3 0.9 þ0.9 9.0 5.1
D Calc to real: Difference between calculated and real costs.
economic data. Despite calculated lower power consumption and energy generation (which is slightly compensated with respect to measured values) electricity income and electricity costs has a deviation of þ1.8% and þ1.3%. Regarding to higher simulated electricity costs it's possible that it depends on individual contracts for energy supply, which are lower than market-standard. As expectable from the good agreement for the technical data, expenses for precipitants and sludge disposal deviated with 0.9% and þ0.9%, respectively only slightly from the real expenses. The deviations of sewer cleaning costs and sewer repair costs are 9.0% and 5.1%. Underestimations of both costs can have different reasons. Either the specific costs or the calculated sewer lengths for cleaning or repair, or both can be too less. With regard to the idealized approaches for aging, rehabilitation and cleaning demand, the estimations of CS of the sewer system and the not calibrated cohort survival functions, the accuracy is satisfying. The annual cost structure (Fig. 4) shows the change in the respective cost categories. In the simulation total costs increased by þ9.2%. For a better comparison, percentages of real and calculated changes and percentage of deviation between real and calculated costs are shown in Table 4. Personnel costs were simulated precisely, the difference between real and calculated values is 360 V (0.03%). By taking into account of contrary price developments at electricity price and precipitation price the development of operating costs are well described, the deviation is 16,881 V (3.84%). Despite minor deviations at sludge disposal costs, sewer cleaning costs and sewer repair costs the largest deviation is at purchased services with 48,698 V (3.79%). The reason may be the dependency of
57
D Calc to real (V)
(%)
Personnel Operating Purchased services Interest Other expenses Depreciation
360 16,881 48,698 þ39,637 þ2071 þ1446
0.03 3.84 3.79 þ5.84 þ0.48 þ0.08
Total costs
22,785
0.35
D Calc to real: Difference between calculated and real costs.
Table 5 Differences between real and calculated amount charge under considering of the accounted reversals. Item
Amount Charge (V m3)
Calculated volume charge Reversal of subsides and contributions
þ2.94 0.35
Adjusted volume charge
þ2.59
rest expenses of cost category on p.e. Expenses do not actually decrease in the same dependency on p.e. as in the model. The largest percentage deviation is at interest with þ5.84% (þ39,637 V). A reason for the deviation can be too high interest rates. If applicable, the utility received loans at better conditions. Calculated cost categories other expenses and depreciation has a high accuracy with a deviation of þ2071 V (þ0.48%) and þ1446 V (þ0.08%). Due to the fact that deviations at depreciation are low, it can be assumed that specific prices for sewer rehabilitation and renewal or the cost functions for the reinvestment of the WWTP are sufficiently precise. Resulting on larger deviations revise each other a minor deviation at total costs of 0.35% (22,785 V) occurs. Usually, rising expenses must lead to increasing basic or volume charge. For a simplified consideration and a better comparison of the results the basic charge is set to constant. In this way, changes in costs will be allocated to the volume charge. The resulting cost structure from the simulation and changing volume charges are depicted in Fig. 4. Within the simulated period, 2001 to 2014 a required increase by þ0.34 V from 2.60 V to 2.94 V (13.3%) was calculated. But, these results still have to be adjusted with respect
Fig. 4. Simulated annual cost structure (left) and simulated annual volume charge (right).
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
58
to accounted reversal of subsidies and contribution (see Table 5). Compared to the real volume charge of 2014 the prognosis just deviates by 1 cent m3. Multiplied with the wastewater volume, the resulting proceeds via the charge are underestimated by 22,785 V. Considering the total utilities revenues statement of 2014, the value corresponds to an underestimation of 0.4%. 3.3. Application for forecasting the development till 2050 After successful validation, the model is exemplarily applied to forecast the development of the utility starting from 2014 until to 2050. For this, technical and economic data as well as the given basic and volume charge are used. Conditions for this period are defined by the wastewater utility. The following assumptions were made:
Condition rates: Industrial growth rate: Population rate: Sewer development rate: Sewer rehabilitation rate: Sewer cleaning rate:
±0.0% 0.5% ±0.0% þ2.0% þ5.3%
Economic rates: Consumer cost rate: Construction costs: Electricity price: Grid feed-in tariff: Personnel cost rate: Sewer cleaning costs: Sewer rehabilitation costs: Sludge disposal costs:
þ2.0% þ2.0% þ2.0% ±0.0% þ1.5% þ2.0% þ2.0% þ2.0%
On the basis of the defined conditions, the resulting development of expenses at each cost category is illustrated in Fig. 5. With regard to the development of personnel costs of þ1.5% per year and the same number of employees, personnel costs would rise by þ793,862 V (60.0%) until the end of the period. As a result of decreasing population, electricity consumption of sewage treatment plants would reduce by 18.2%, electricity generation by 2.4% and precipitation consumption by 11.6%. However, in spite of decreasing operating materials, due to inflation expenses of the cost category would increase by þ67,618 V (þ16.0%). The cost of purchased services would increase by þ517,571 V (38.1%), including þ198,604 V (51.6%) for disposal of sewage sludge, þ62,026 V (þ775.3%) for sewer repairs and further þ5973 V (þ79.6%) for sewer cleaning. Regarding sewage sludge, the mass would decrease by 11.9%, but caused by inflation and the consideration that in future sludge income will (for legal reasons) probably be incinerated instead of agricultural valorization the costs for sewage sludge disposing would increase. The expenses of
Fig. 6. Simulated annual volume charge.
sewer repairs would increase despite an annually renovation rate of 2%. This is due to the special homogeneous age structure of the sewer system, which leads despite the high investments in rehabilitation and renovation to deteriorating sewer conditions. Furthermore, expenses for interest would rise by þ236,977 V (þ32.3%), other expenses decreased by 47,442 V (9.1%) and depreciation would rise by þ98,529 V (þ5.4%). The charge resulting from the sum of expenses, reduced by resolutions, is shown in Fig. 6. On the basis of the given conditions the volume charge would increase by þ1.23 V (þ47.4%) from 2.60 V to 3.83 V. By the reversal from the residual value of contributions in 2021 the calculated charge jumps. The charge reduction in 2023 as well as a stagnation in 2045 result from the reinvestments of WWTP, which are dimensioned to reduced p.e. at time of investment. 4. Conclusion The developed simplified process-based model for technical and economic quantification can close a gap between desired technical and economic forecasting based on process based relationships and the necessity to simplify those forecasts for reasons of applicability
Fig. 5. Simulated annual cost structure from 2014 to 2050.
€nckner / Journal of Environmental Management 207 (2018) 51e59 M. Giessler, J. Tra
and data availability. The model has been developed in close cooperation with several utilities and addresses their requirements. The validation with historical data provided by an exemplary utility (time series of 13 years) has shown that the results reflect well both, technical key data and derived current costs and charges. While the calculation approaches for comparable technical systems are transferrable (which is the case for many European regions), the cost accounting model may need adaptations when transferring it to different accounting rules. The accuracy of the model can be improved by case specific definition of calculation parameters. This applies namely for probably cost sensitive calculations like the cohort survival functions in module Sewer. In the case given, parameter for this function literature value had to be used (Jansen, 2007), due to missing data. The process oriented setup of the model enables its user to assess widely varying development scenarios and technical as well as organizational management options showing differentiated causal relationships. Acknowledgements The study was funded by a state graduate scholarship of the state Mecklenburg Vorpommern, Germany. We would like to thank all stakeholders and ministries for their cooperation and support during the investigations. References €sser e AbwAG, 2005. Gesetz über Abgaben für das Einleiten von Abwasser in Gewa Abwasserabgabengesetz (Law on charges for the discharge of wastewater into waters e Wastewater Levy Act). Germany, Version of the Announcement from 18.01.2005. Ana, E.V., Bauwens, W., 2010. Modeling the structural deterioration of urban drainage pipes: the state-of-the-art in statistical methods. Urban Water J. 17, 47e59. ATV - DVWK - A 131, 2000. Bemessung von einstufigen Belebungsanlagen (Dimensioning of single-stage activated sludge plants). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. €rpern und RotationATV - DVWK - A 281, 2001. Bemessung von Tropfko € rpern (Dimensioning of trickling filters and rotating biological constauchko tactor). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. BDEW, 2015. Industriestrompreise: Energieinfo - Ausnahmeregelungen bei Energiepreis-bestandteilen (Industrial electricity prices: Energy information - Derogations for energy price components). Bundesverband der Energie- und Wasserwirtschaft e. V., Berlin, Germany. BMWi, 2016. EEG in Zahlen: Vergütungen, Differenzkosten und EEG-Umlage 2000 bis 2017 (EEG in numbers: Remuneration, differential costs and EEG allocation 2000 to 2017). Bundesministerium für Wirtschaft und Energie, Berlin, Germany. Castellet, L., Molinios-Senante, M., 2016. Efficiency assessment of wastewater treatment plants: a data envelopment analysis approach integrating technical, economic, and environmental issues. J. Environ. Manage. 167, 160e166. €tze für Bemessung, Bau und Betrieb von AbwasserteiDWA-A 201, 2005. Grundsa chanlagen (Principles for dimensioning, construction and operation of wastewater treatment ponds). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. DWA-M 149-3, 2015. Zustandserfassung und -beurteilung von Entw€ asser€uden e Teil 3: Beurteilung nach optischer ungssystemen außerhalb von Geba Inspektion (Conditions and Assessment of Drain and Sewer Systems Outside Buildings - Assessment by visual inspection). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. €sserung (Mechanical sludge dewDWA-M 366, 2000. Maschinelle Schlammentwa atering). DWA-Regulations, German Association for Water, Waste Water and Waste, Hennef, Germany. Ferreira da Cruz, N., Carvalho, P., Marques, R.C., 2013. Disentangling the cost
59
efficiency of jointly provided water and wastewater services. Util Policy 24, 70e77. Gretzschel, O., Schmitt, T.G., Hansen, J., Siekmann, K., Jakob, J., 2014. Sludge digestion instead of aerobic stabilisation - a cost benefit analysis based on experiences in Germany. Water Sci. Technol. 69, 430e437. Haider, H., Sadiq, R., Tesfamariam, S., 2016. Intra-utility performance management model (In-UPM) for the sustainability of small to medium sized water utilities: conceptualization to development. J Clean Prod. 133, 777e794. Henze, M., Gujer, W., Mino, T., van Loosdrecht, M.C.M., 2000. Activated Sludge Models ASM1, ASM2, ASM2d and ASM3. IWA Task Group on Mathematical Modeling for Design and Operation of Biological Wastewater Treatment. IWA Publishing, London, England. Hernandez-Sancho, F., Molinos-Senante, M., Sala-Garrido, R., 2011. Cost modeling for wastewater treatment processes. Desalination 268, 1e5. Herz, R., 1996. Aging processes and rehabilitation needs of drinking water distribution networks. J. Water SRT Aqua 45, 221e231. Imhoff, K., Imhoff, K.R., 2007. Taschenbuch der Stadtentw€ asserung (Paperback of urban drainage). Oldenbourg-Verlag, Munich, Germany. Jansen, K., 2007. AQUA-WertMin Program Manual. Expert Office for Sewer, Klienbittersdorf, Germany. KAG M-V, 2005. Kommunalabgabengesetz e KAG M e V (State Municipal Charges Act of the state Mecklenburg e Vorpommern). State MecklenburgVorpommern, Germany, Version of the Announcement from 12.04.2005. Kleiner, Y., Rajani, B., 2001. Comprehensive review of structural deterioration of water mains: statistical models. Urban Water 3, 131e150. Kley, G., Caradot, N., 2013. Review of Sewer Deterioration Models e Project Acronym: SEMA. Report from Kompetenzzentrum Wasser Berlin gGmbH. Department Surface Water, Berlin, Germany. Le Gat, Y., 2008. Modelling the deterioration process of drainage pipelines. Urban Water J. 5, 97e106. LUNG, 2015. Kommunale Abwasserbeseitigung in Mecklenburg-Vorpommern: €ß Richtlinie 91/271/EWG (Sewage disposal in Lagebericht 2015-Bericht gema Mecklenburg-Vorpommern: Status report 2015 e Report according to Directive 91/271/EEC). Landesamt für Umwelt, Naturschutz und Geologie MecklenburgVorpommern, Güstrow, Germany. MLUV, 2009. Zustand der Abwasserkanalisation in Mecklenburg-Vorpommern (Conditon state of sewer networks in Mecklenburg-Vorpommern). Ministerium für Landwirtschaft, Umwelt und Verbraucherschutz Mecklenburg-Vorpommern, Schwerin, Germany. Rajani, B., Kleiner, Y., 2001. Comprehensive review of structural deterioration of water mains: physically based models. Urban Water 3, 151e164. Rehan, R., Knight, M.A., Unger, A.J.A., Haas, C.T., 2014. Financially sustainable management strategies for urban wastewater collection infrastructure e development of a system dynamics model. Tunn. Undergr. Sp. Technol. 39, 116e129. Ruiz-Rosa, I., García-Rodríguez, F.J., Mendoza-Jimenez, J., 2016. Development and application of a cost management model for wastewater treatment and reuse processes. J. Clean Prod. 113, 299e310. Silva, C., Saldanha Matos, J., Rosa, M.J., 2016. Performance indicators and indices of sludge management in urban wastewater treatment plants. J. Environ. Manage. 184, 307e317. €lkerStatA MV, 2013. Aktualisierte 4. Landesprognose - Basisjahr 2010: Bevo ungsentwicklung des Landes Mecklenburg e Vorpommern sowie 48 der €dte und Landkreise bis 2030 nach Altersgruppen (Updated 4th kreisfreien Sta country forecast - base year 2010: Population change of Mecklenburg e Vorpommern as well as 48 of the independent cities and counties until 2030 by age groups). Statistical reports. Bd. A 1831 2012 01, State Statistical Office of Mecklenburg e Vorpommern, Schwerin, Germany. StatBA, 2016. Preisindizes für die Bauwirtschaft (Price indices for the construction industry). Statitisches Bundesamt, Fachserie 17 Reihe 4, Artikelnummer: 2170400163244, Wiesbaden, Germany. StatBA, 2017. Preise: Verbraucherpreisindizes für Deutschland - Lange Reihen ab 1948 (Prices: Consumer price indices for Germany - Long series from 1948). Statitisches Bundesamt, Artikelnummer: 5611103161124, Wiesbaden, Germany. €nckner, J., Franz, T., Frehmann, T., Jathe, R., Obermayer, A., Winkler, U., 2014. Tra Wirtschaftliche Auswirkungen sich überlagernder Entwicklungstrends auf Abwasserentsorgungsunternehmen (Economic effects of overlapping trends on wastewater utilities). KA Korrespondenz Abwasser 61, 793e801. Yang, J., 2004. Road Crack Condition Performance Modeling Using Recurrent Markov Chains and Artificial Neural Networks. Doctoral Dissertation. Department of Civil and Environmental Engineering, College of Engineering,University of South Florida.