ENERWATER - A standard method for assessing and improving the energy efficiency of wastewater treatment plants

1 ENERWATER - A standard method for assessing and improving the energy efficiency of wastewater treatment plants Longo S.a, *, 1 , Mauricio-Iglesias M. a, Soares A. b, Campo P. b, Fatone F. c, Eusebi A. L. c, Akkersdijk E. d, Stefani L. e, Hospido A.a a Department of Chemical Engineering, Institute of Technology, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain b Cranfield Water Science Institute, Cranfield University, Cranfield, Bedfordshire MK43 0AL, UK c Department of Materials, Environmental and City Planning Science and Engineering, Faculty of Engineering, Polytechnic University of Marche, Ancona, Italy c Aggerverband, FB Abwasserbehandlung Sonnenstraße 40, 51645 Gummersbach, Germany e ETRA, via del Telarolo 9, 35013 Cittadella, Italy * Corresponding author (e-mail: stefano[email protected]; [email protected]) Abstract This paper describes the first methodology specifically tailored to estimate energy efficiency at wastewater treatment plants (WWTPs). Inspired by the cycle of continuous improvement, the method i) precisely defines the concept of energy efficiency in WWTPs, ii) proposes systematic and comparable ways to measure it, and iii) allows benchmarking and diagnosing energy hotspots. The methodology delivers an aggregated measure of the WWTP energy efficiency defined as the Water Treatment Energy Index, a single energy label that uses universally known illustrations enabling wide communication of standardized information on the WWTP energy status. The accuracy, reproducibility and generality of the methodology were validated by a widespread energy benchmarking method, and a case study is presented to show its capabilities. By promoting dialogue towards the 1 Present address: Gruppo Hera S.p.A, 40127 Bologna, Italy. 2 creation of a specific European Standard, the actions accomplished by the H2020 Coordination Support Action ENERWATER should positively contribute to improving the exchange of information on energy saving actions and results between wastewater utilities and towards other stakeholders. Keywords WWTP; key performance indicators (KPI); benchmarking; label; diagnosis Nomenclature BOD Biochemical oxygen demand CED Cumulative energy demand CFU Colony-forming unit COD Chemical oxygen demand DEA Data envelopment analysis DS Decision support EPI Energy performance indicators EU European Union GPP Green public procurement KPI Key performance indicator kW Kilowatt N Nitrogen PE Person equivalent P Phosphorus RA Rapid audit TN Total nitrogen TP Total phosphorus TSS Total suspended solids 3 UV Ultraviolet WWTP Wastewater treatment plant WTEI Water treatment energy index 1. Introduction Water and energy are highly interconnected. Water is needed for most stages of energy production and transmission, and energy is crucial for the provision and treatment of water. This fundamental resource relationship is called the water-energy nexus [1]. The higher the water use by end users, the higher the energy use, and then the higher the water use for energy production, resulting in a feedback loop and ultimately in higher carbon emissions. The increase in carbon emissions contributes to climate change [2], which negatively impacts the availability of water and energy, and shortages in one resource can directly affect the availability of the other. With both water and energy needs set to increase [3], it has become ever more important to understand the linkages between the two, to anticipate future stress points and implement policies, technologies and practices that soundly address the associated risks. Of the energy consumed along the urban water cycle, the largest amount is used for wastewater treatment, in the form of electricity, in developed countries [4]. To counterbalance the increasing trend in energy intensity of wastewater treatment processes, energy efficiency improvement is the only option as effluent quality needs to be ensured [5]. Any energy policy in the wastewater sector should lead to reduced energy consumption without compromising public health and environment. In practise, such a policy implies i) using less energy to treat the same amount of wastewater or ii) treating more wastewater (or more thoroughly) with the same amount of energy. Both cases require wastewater treatment to become more energy efficient. With the transposition of the Energy Efficiency Directive 2012/27/EU [6], carrying out energy audit at wastewater treatment plants (WWTPs) has evolved from convenient to an obligation for a significant 4 part of European water utilities, i.e. those with more than 250 employees and with annual trading volume greater than €50 million or whose annual balance sheet exceeds €43 million. However, the Directive as well as its transposition into national legislation by the Member States lacks sufficient detail for a clear and consistent implementation [7]. First, the concept of energy efficiency for WWTPs is not clearly defined: although the Directive defines energy efficiency as "the relationship between the production of service, good or energy and energy demand", the service provided by WWTPs, i.e. "cleaning wastewater", must be specified in quantitative objectives/functions such as "eliminate organic carbon", "eliminate nitrogen", "eliminate solids" or "eliminate pathogens", etc., depending on, e.g., the quality of the effluent wastewater and the location of the discharging point [8]. Second, WWTPs are intrinsically characterized by having heterogeneous layouts, which makes comparisons not trivial. Indeed, the treatment processes are organized in different unit operations grouped together to provide various levels of treatment known as preliminary, primary, secondary, tertiary and sludge treatment [9]. Depending on economic and environmental criteria, WWTPs are composed by different combinations of treatment levels. This heterogeneity has led some scholars to think that WWTPs benchmarking is unfeasible, given that each plant is different [10]. Measurement is however the first step that leads to information gathering, control and eventually improvement. Therefore, any successful WWTPs energy benchmarking system should be capable to adapt to the different WWTPs layouts and process schemes commonly used in the wastewater sector. In a previous publication [11], we revised existing literature on WWTP energy-use performance and of the state-of-the-art methods for WWTP energy benchmarking, eventually identifying the need of a standardised method. This paper intends to fill that gap by presenting a methodology for carrying out energy benchmarking and diagnosis of energy efficiency of WWTPs. Besides, we show how the absence of systematic methodologies for plant wide evaluation of energy efficiency and the lack of a procedure for benchmarking WWTPs energy efficiency represented a major obstacle to improve WWTPs energy performance. While energy efficiency guidelines and measures are sometimes available for specific equipment, such as blowers [12] or pumps [13], there is no clear way to determine how 5 well these components operate at plant-wide level. Furthermore, decision-makers require tailored energy-related metrics in order to communicate the status quo adequately with other stakeholders [14]. To answer to the European normative pressure and avoid economically wasteful energy policies the need for standardization in the evaluation and comparison of WWTPs energy efficiency appears even more necessary. Based on both a solid theoretical foundation and feedback from the wastewater sector stakeholders, we propose here a methodology as an energy efficiency benchmarking framework. This study therefore focuses on the development of a structured and systematic method for assessing and improving the energy efficiency in WWTPs. The key novelty of the ENERWATER methodology lies on its output – the Water Treatment Energy Index (WTEI) - a single energy label that uses universally known illustrations, to widely communicate standardized information on the WWTP energy status. The methodology here presented is relevant to anyone interested in WWTPs energy efficiency as, for the first time, it provides engineers, wastewater operators and decision-makers a method to obtain standardized and comparable efficiency information. In particular by offering guidelines on how to define energy efficiency of WWTPs and identifying the sources of energy misuse, the outcomes of this article are expected to move WWTPs towards increasing energy efficiency. Section 2 provides theoretical background on energy efficiency benchmarking methods focusing on key challenges. Section 3 includes a description of the methodology, together with a step-by-step demonstration of a selected case study. The robustness of the efficiency estimation process is then validated in Section 4, while the necessity and utility of the methodology as well as current limitations and future outlook are discussed in Section 5. Finally, Section 6 offers concluding observations. 2. Literature review In the last decade, a number of benchmarking tools have been developed to estimate energy (in)efficiency in industrial systems. A traditional way to overcome some of the difficulties of making comparison is to use Key Performance Indicators (KPIs) [15], which reflect the purpose of the facility 6 under comparison. A KPI is often a ratio of an input and an output and is usually employed and obtained by simply normalizing the energy use based on the unit activity or service provided [16]. KPIs are often used to monitor energy performance in several industrial applications, e.g. from subway stations [17] to compressed air systems [18]. In the wastewater sector, KPIs have been used to give a general overview of the energy performance of WWTPs [19]. Although such simple normalization is relatively inexpensive to apply, is not data intensive and is easy to implement and understand, the downside it its very limited in scope as KPIs involves only partial evaluations, which is an important constraint [11]. So, a single KPI may not fully reflect the purpose of the plant. A WWTP, from a functionality point of view, could have multiple outputs, e.g. removing chemical oxygen demand (COD), nitrogen, phosphorus, and pathogens, or producing energy or material like biogas and fertilizers. In this regard, a proper measure of WWTP energy efficiency should reflect a multidimensional concept (i.e. taking into account for the different functions of the plant). Although benchmarking methods based on multiple KPIs have been discussed, such as by Fraia et al. [20], some sort of weighting between different KPIs would be necessary. Otherwise, it can be difficult to interpret the results of different indicators, since trade-offs exist for WWTPs at different stages of their lifecycles. In order to overcome the previous limitations, Data Envelopment Analysis (DEA) represents an attractive tool for performance assessment and, focusing on the last 10 years, there are growing number of studies adopted DEA in energy efficiency analysis. Thanks to its ability to handle multiple inputs and outputs, DEA models have been used to evaluate energy and environmental performance of complex systems such as chemical processes [21], industrial gases facilities [22], service sector [23], and including water [24] and wastewater treatment facilities [25]. Although DEA has great potential for energy efficiency evaluation of WWTPs, it is hardly extensible at international level as a standard tool. In effect, including a new WWTP requires solving the DEA model again for the whole set of observations, with potential changes in the established ranking if the new plant in the set moves away the frontier. 7 A third category of benchmarking methods is based on regression analysis and it is called parametric approach. Regression models describe the relationship between energy use of a system and predictor variables influencing energy use, including characteristics and external factors. A consequence of using parametric approaches is that the residuals (i.e. the difference between the energy use predicted by the model and the actual energy use) are treated as a measure of efficiency [26]. The parametric approach is widely used in building energy efficiency applications [27], and it is especially employed for exploring the effects of influencing factors on energy efficiency [28] and identifying its determinants [29]. Furthermore, using specific Stochastic Frontier Analysis models and panel data it is possible to distinguish between persistent and transient inefficiency [30], which is particularly useful in the wastewater sector to complete appropriate energy efficiency diagnosis of WWTPs [31]. However, it is not straightforward and an important source of debate to decide which factors are legitimate uncontrollable influences on performance, and hence to be included in the regression model, and which are within the control of the management. For example, structural differences such as plant size and load factor are compensated in the Energy Star method for WWTPs developed by the US Environmental Protection Agency [32], while they may originate from inefficient plant design. This discussion suggests that controversies may arise from the use of parametric approaches when it comes to standardization applications, while any standard method should be universally applied independently of the stakeholder who is employing it. In the last decade, efforts in the industry have been targeted to achieve energy efficiency at WWTPs and energy benchmarking systems at WWTPs have become common practice in some countries. Good examples are the detailed energy management systems developed in Germany and Austria [5]. Those approaches are however hardly extensible at international level due to the fact that, using load-specific energy use stated as kWh/PE·y, where PE stands for the Person Equivalent, they assume that concentrations in the influent and effluent (e.g. solids, organic matter, nitrogen, phosphorus etc.) do not vary significantly between WWTPs, hence restricting the application of these approaches to homogenous geographical area with similar effluent quality requirements. 8 The review of pertinent literature reveals that both academia and industry still lack standard approaches and tools to quantify energy efficiency, able to accurately define WWTPs energy efficiency, to adapt to different plant layouts, possibility of including energy produced onsite, having good geographical coverage at European level, and being of easy communication by an aggregated indicator that reflects the complexity of a WWTP. Based on the previous discussed limitations of existing energy efficiency benchmarking approaches, the main goal of the ENERWATER methodology is to contribute to development of the standardised EU energy methodology and labelling in WWTPs. This study is built upon the results of the ENERWATER 2 project, a three-year Coordinated Support Action within the Horizon 2020 program. 3. The ENERWATER methodology The methodology developed in the framework of the ENERWATER project 3 aims to systematically determine the energy efficiency of a particular WWTP expressed by the WTEI. The methodology includes the definition of WWTPs typologies, the classification of facilities accordingly, the identification of levels of treatment (stages), the identification of the correspondent KPIs, their aggregation into a composite index (i.e. WTEI), and its labelling for an easy and straightforward communication. 3.1 General considerations 3.1.1 Rapid Audit and Decision Support versions The methodology can be applied in two different ways according to the following goals: 2 The reader is referred to www.enerwater.eu for further information. 3 All public deliverables are available in the project website at the following link www.enerwater.eu/downloaddocumentation. 9 – The Rapid Audit (RA) method leads to quick estimation of the WTEI based on existing information, such as historical energy use data along with influent and effluent quality values obtained by routine analyses. By doing so, the aim is to obtain a WWTP energy benchmark, a rapid tool to compare a given WWTP performance with other plants and ascertain the need for a detailed monitoring campaign. – The Decision Support (DS) method requires intensive monitoring of energy use and water quality parameters to provide an accurate and detailed calculation of the WTEI for each WWTP stage as well as its overall value for the plant. By doing so, the aim is to serve as a diagnosis of the functions/equipment in order to individuate the origin of inefficiency and develop targeted energy saving strategies. Both methodologies are structured in a similar way but require inputs with a different level of detail (Fig. 1). In both cases, all measured data can be reported as daily, monthly or yearly averages, being 3 years the recommended time period for data gathering to account for seasonal variability associated with the human activities and the seasonal rainfall. Due to the variable influent behaviour, the pollution load to be treated is continuously changing, and consequently, so are the energy and chemical requirements for the treatment [33]. To sum up the procedures, first the type of WWTP is established according to its functions; then, energy consumption and other measurements (flowrate, pollutant concentrations, etc.) are combined to obtain the relevant KPIs, which are then normalised and weighted to obtain the WTEI. Finally, the WTEI is presented as an energy label to provide all stakeholders with standardized information and facilitate dissemination of WWTPs’ the energy efficiency. 16 Figure 3. Workflow for the Water Treatment Energy Index calculation. 3.2.1 Step 1: Estimation Energy consumption data Historical data on the energy consumed at the WWTP need to be available, including electricity and other fuels such as diesel, natural gas etc. Total WWTPs electricity consumption can be obtained by consulting electricity bills (only for RA), meter readings or existing on-line meters. Likewise, the disaggregated electricity consumption (required for DS) can be measured or estimated, combining the rated power of the electrical motor in kilowatt (kW) and the working hours in a year to provide an estimation of kWh used in each stage per unit of time. If other energy sources are used, for example to drive generators to produce electricity, they need to be quantified and converted into kWh per unit of time (Table 3) to calculate the total energy consumption (Eq. 1). s Step 2 Step 1 Step 3Step 4Step 5 Key Performance Indicators (KPI) estimation Ranking Energy Performance Indicators (EPIs) EPIs Aggregation Weights selection Comparison with distribution functions WTEI calculation Data collection and KPIs calculation Compare the value of the KPIs with distribution functions and obtain percentile of each KPIs (EPIs) Choose the weights for the selected KPIs from database of WWTPs energy consumption data Aggregate the EPIs into a single WTEI through a weighted sum Assign label corresponding to the value of the WTEI 17 𝐸1: 𝐸𝑛𝑒𝑟𝑔𝑦 𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑑 = 𝐸𝑝𝑉1 + 𝐸𝑝𝑉2+ 𝐸𝑝𝑉3+ 𝐸𝑝𝑉4 [𝑘𝑊ℎ 𝑦𝑒𝑎𝑟] (1) Where, EpVi is energy consumed as electric energy (V1), diesel (V2), natural gas (V3) and biogas (V4). Table 3. Energy carriers and associated conversion factors applied by the ENERWATER methodology. Energy carrier Conversion factors Abbr. Equations to estimate specific power consumption Electric energy in kWh 1 ( kWhkWh ⁄) V1 𝐸𝑝𝑉1 = 𝑃 × 𝑇  𝑈𝑠𝑒 𝐹𝑎𝑐𝑡𝑜𝑟 Diesel in kg 11.87 ( kWhkg ⁄) V2 𝐸𝑝𝑉2 = 𝑒𝑞𝑢𝑖𝑝𝑚𝑒𝑛𝑡 𝑢𝑠𝑎𝑔𝑒 /𝑦𝑒𝑎𝑟 𝑥 𝑢𝑠𝑎𝑔𝑒 𝑡𝑖𝑚𝑒 (ℎ)  11.87  𝑑𝑖𝑒𝑠𝑒𝑙 𝑢𝑠𝑒𝑑 [𝑘𝑔 /ℎ]  𝛽𝑒𝑙∗ Natural gas in Ncm‡ 9.94 (kWh Ncm ⁄ ) V3 I) 𝐸𝑝𝑉3 = 𝐺𝑎𝑠 𝑖𝑛 𝑐𝑜𝑚𝑏𝑖𝑛𝑒𝑑 ℎ𝑒𝑎𝑡 𝑎𝑛𝑑 𝑝𝑜𝑤𝑒𝑟 𝑒𝑛𝑔𝑖𝑛𝑒 𝑘𝑊ℎ𝑒𝑙 = (𝑁𝑐𝑚/𝑦)  9.94  𝛽𝑒𝑙∗ 𝑘𝑊ℎ𝑡ℎ = (𝑁𝑐𝑚/𝑦)  9.94  (1 − 𝛽𝑒𝑙))  𝛽𝑡ℎ† II) 𝐸𝑝𝑉3 = 𝐺𝑎𝑠 𝑢𝑠𝑒𝑑 𝑓𝑜𝑟 ℎ𝑒𝑎𝑡𝑖𝑛𝑔 𝑜𝑛𝑙𝑦 𝑘𝑊ℎ𝑡ℎ = 𝑁𝑐𝑚/𝑦  9.94  𝛽𝑡ℎ† Biogas in Ncm‡ 9.94 𝑥 𝑁𝐺𝐶 (kWh Ncm ⁄ ) where NGC is the natural gas content in the biogas (vol/vol) V4 I) 𝐸𝑝𝑉4 = 𝑏𝑖𝑜𝑔𝑎𝑠 𝑖𝑛 𝑐𝑜𝑚𝑏𝑖𝑛𝑒𝑑 ℎ𝑒𝑎𝑡 𝑎𝑛𝑑 𝑝𝑜𝑤𝑒𝑟 𝑒𝑛𝑔𝑖𝑛𝑒 𝑘𝑊ℎ𝑒𝑙 = (𝑁𝑐𝑚 𝑦)  9.94  𝑁𝐺𝐶 𝑥 𝛽𝑒𝑙∗ 𝑘𝑊ℎ𝑡ℎ = (𝑁𝑐𝑚 𝑦)  9.94  𝑁𝐺𝐶 𝑥 (1 − 𝛽𝑒𝑙))  𝛽𝑡ℎ† II) 𝐸𝑝𝑉3 = 𝐵𝑖𝑜𝑔𝑎𝑠 𝑢𝑠𝑒𝑑 𝑓𝑜𝑟 ℎ𝑒𝑎𝑡𝑖𝑛𝑔 𝑜𝑛𝑙𝑦 𝑘𝑊ℎ𝑡ℎ = 𝑁𝑐𝑚/𝑦  9.94  𝛽𝑡ℎ† * Typical efficiency = 0.40 for electricity generation; † typical efficiency = 0.85 for heat production and recovery ‡ Ncm = normal cubic meters. Normal conditions (0°C, atmospheric pressure) 18 Chemical energy consumption In some WWTPs chemicals such as iron sulphate or iron chloride are added to the wastewater to remove pollutants such as phosphorus. Other chemicals that are frequently used in WWTPs include alum, polyelectrolyte, acetate, methanol etc. Hence, the use of chemicals and their specific dosage can impact the pollutants’ removal efficiency of WWTPs and replace, to a certain extent, the use of energy. The trade-off between energy and chemicals use was tackled in the ENERWATER methodology by using the Cumulative Energy Demand (CED) method developed by Frischknecht et al. [39], which is a widely used indicator for environmental impact evaluations [40]. It reports the direct and indirect consumption of energy necessary to obtain a product or service by computing the equivalent of primary energy consumption in the product chain or the energy consumed in a certain system over its entire lifecycle. Chemical energy consumptions for the main chemicals used during wastewater treatment are given in Table S1 of Supplementary Material. Equation 2 represents the formula used for estimating the chemical energy consumption due to the chemicals. 𝐸2: 𝐶ℎ𝑒𝑚𝑖𝑐𝑎𝑙 𝑒𝑛𝑒𝑟𝑔𝑦 𝑐𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 = ∑𝑐𝑒𝑐𝑖×𝑀𝑖 L i=A [𝑘𝑊ℎ 𝑦𝑒𝑎𝑟] (2) Where, A to L are the chemicals used in the WWTP, 𝑀𝑖 is the mass (in kg) consumed of each chemical and 𝑐𝑒𝑐𝑖 is the specific chemical energy consumption (in kWh/kg) for all chemicals used in the WWTP from A to L: A - Acetic acid; B - Aluminium sulphate; C - Iron(III) chloride; D - Iron(III) sulphate; E - Iron(II) sulphate, F - Methanol; G - Peracetic acid; H - Poly-Aluminium-Chloride; I - Polyelectrolyte; L - Sodium hypochlorite. WWTPs producing energy producing and sludge imports Wastewater treatment plants can have a range of technologies that produce energy/electricity on site, such as anaerobic digestion (of sludge, imported sludge, other wastes, etc.), hydraulic-power, wind 19 turbines, solar panels, fuel-cells, etc. The generation of electricity in the WWTP can (partially) offset the energy demand of the facilities and should be accounted by using Equation 3. 𝐸3∶𝐸𝑛𝑒𝑟𝑔𝑦 𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑑 𝑎𝑡 𝑊𝑊𝑇𝑃 = ∑𝑖 𝐿 𝑖=𝐴 [𝑘𝑊ℎ 𝑦𝑒𝑎𝑟] (3) Where, 𝐴 to 𝐿 are the types of energy produced in the WWTP: A – biogas (kWh/year); B - hydraulicpower (kWh/year); C - wind turbines (kWh/year); D - solar panels (kWh/year); E – fuel-cells (kWh/year); F-L – other (kWh/year). When considering anaerobic digestion, many WWTPs act as sludge treatment centres receiving sludge from nearby sites. This imported sludge is often mixed with the sludge produced at the WWTP for further treatment such as dewatering, anaerobic digestion etc., raising significant shares in some WWTPs (up to 2-fold the sludge produced on site). As a result, the volume of sludge imports, respective total suspended solids as well as an estimation of the energy consumed and produced for its treatment, needs to be taken into consideration (Equation 4). 𝐸4: 𝐸𝑛𝑒𝑟𝑔𝑦 𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑑 𝑎𝑛𝑑 𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑑 𝑏𝑦 𝑠𝑙𝑢𝑑𝑔𝑒 𝑖𝑚𝑝𝑜𝑟𝑡𝑠 = 𝐸𝑛𝑒𝑟𝑔𝑦 𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑑 − 𝐸𝑛𝑒𝑟𝑔𝑦 𝑐𝑜𝑛𝑢𝑚𝑒𝑑 (𝑏𝑦 𝑠𝑙𝑢𝑑𝑔𝑒 𝑖𝑚𝑝𝑜𝑟𝑡𝑠)[𝑘𝑊ℎ 𝑦𝑒𝑎𝑟]5 (4) Total gross and net energy consumption estimation The gross and net energy consumed can be estimated by combining the results from Equations 1-4 as well as sludge imports (Equation 5 and 6, respectively). Gross and net energy consumptions are used as input to estimate each KPI. 5 It can be assumed that energy produced and consumed by sludge imports = 𝐸3 x (sludge imports/total amount of sludge). In case 𝐸4 would be negative (i.e. the energy consumed by sludge imports is higher than the energy that it produces) it should be considered equal to zero given that sludge deriving from other plants is out of the boundaries of the plant and it is not considered a plant function. 20 𝐺𝑟𝑜𝑠𝑠 𝑒𝑛𝑒𝑟𝑔𝑦 𝑐𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 = 𝐸1+𝐸2 [𝑘𝑊ℎ 𝑦𝑒𝑎𝑟] (5) 𝑁𝑒𝑡 𝑒𝑛𝑒𝑟𝑔𝑦 𝑐𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 =𝐸1+𝐸2 – 𝐸3+ 𝐸4 [𝑘𝑊ℎ 𝑦𝑒𝑎𝑟] (6) Identification of KPIs and calculation of its reference values For the RA methodology it is recommended that different KPIs are considered taking in consideration influent and effluent data (i.e. routine analysis normal available). This information can be obtained from the flowrate measurements taken through online flow meters or similar, or the information taken from the WWTP design sheets. For the DS methodology it is recommended that different KPIs are considered by means of composite or grab samples to account for the key pollutants removed at the different stages of the process (i.e. by detailed sampling campaign is required). Suggested KPIs for application of RA and DS methodology are given in Table 4. Table 4. Identification of KPIs. Plant function Stage Parameter Rapid Audit Decision Support Pumping S1 Flow kWh/m3 Removal of suspended solids* S2 Total Suspended Solids (TSS) - kWh/kg TSSrem Removal of organic matter S3 Chemical Oxygen Demand (COD) kWh/kg TPErem 6 Removal of nitrogen S3 Total Nitrogen (TN) Removal of phosphorus S3 Total Phosphorus (TP) Removal of pathogens S4 E. coli Colony-forming Unit (UCF) kWh/LogRed*m3 6 𝑇𝑜𝑡𝑎𝑙 𝑃𝑜𝑙𝑙𝑢𝑡𝑖𝑜𝑛 𝐸𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡 (𝑇𝑃𝐸)=𝐶𝑂𝐷 (𝑘𝑔𝐶𝑂𝐷)+20 𝑇𝑁 (𝑘𝑔𝑇𝑁)+100 𝑇𝑃 (𝑘𝑔𝑇𝑃) [35]. 21 Removal of produced sludge* S5 Total Solids (TS) kWh/kg TSproc kWh/kg TSE7 Sludge dewatering S5 Total Solids (TS) *It applies only to RA methodology For the RA, KPIs relate the overall energy consumption (e.g. gross energy consumption). In the DS, KPIs are directly associated with the appropriate stage. In this case, the KPIs are calculated using the specific portion of energy consumption related to its function. Summary statics of database of KPIs for RA and DS methodologies are given in Table 5 and 6, respectively. Table 5. Database of KPIs for overall plant and Rapid Audit methodology. KPI KPI units Average St. Dev. P90 P10 Obs. S1 kWh/m3 0.348 0.445 0.901 0.161 97 S3 kWh/kg TPErem 0.488 0.292 0.731 0.171 87 S4 kWh/(LogRed·m3) 0.058 0.076 0.137 0.030 53 S5 kWh/kg TSproc 2.074 2.165 5.231 0.824 89 Table 6. Database of KPIs for Decision Support methodology. KPI KPI units Average St. Dev. P90 P10 Obs. S1 kWh/m3 0.048 0.039 0.101 0.009 97 S2 kWh/kg TSSrem 0.028 0.030 0.055 0.007 64 S3 kWh/kg TPErem 0.289 0.246 0.519 0.108 87 S4 kWh/(LogRed·m3) 0.030 0.047 0.054 0.010 53 S5 kWh/kg TSE 0.308 0.400 0.577 0.055 89 7 𝑇𝑜𝑡𝑎𝑙 𝑆𝑜𝑙𝑖𝑑 𝐸𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡 (𝑇𝑆𝐸)=𝑇𝑆𝑟𝑒𝑚𝑜𝑣𝑒𝑑 (𝑘𝑔𝑇𝑆)+ 2 𝑇𝑆𝑑𝑒𝑤𝑎𝑡𝑒𝑟𝑒𝑑 (𝑘𝑔𝑇𝑆). Weights are estimated based on own calculations using the ENERWATER dataset [37]. 22 3.2.2 Step 2: Normalization The KPIs are expressed in a variety of units. Hence, there is need to express them on a common basis. Normalization is done here by comparison with a distribution function, so that the percentiles for each KPI are normalised indicators of performance, here called energy performance indicators (EPI). By comparing the value of the KPIs with the database distribution function a percentile for each KPI is obtained. The percentile is a normalized manner to express the performance of the plant for a given KPI. Each KPI can be normalized by using Eq. 7, which corresponds to Gumbel’s cumulative distribution function with parameters estimated for the population of WWTPs in the benchmark database (Table S2 of Supplementary Material). EPIi=Percentile(%) = exp(−exp(−(KPIi−μ𝑖 σ𝑖)))x100 (7) 3.2.3 Step 3: Weights selection Weighting emphasizes the contribution of a given KPI over others in terms of energy consumption. The particular weights to be applied at ENERWTATER methodology (Table 7) have been estimated based on the average relative contribution of each function/section of the WWTP to the overall energy consumption based on the ENERWATER database (Section 3.1.6), i.e. pumping (stage 1) accounts for almost 12% of the overall energy consumption and the secondary treatment (stage 3) accounts for 54%. The proportions of energy consumption associated with different plant sections (from which the weights have been extrapolated) are in agreement with those available in the literature. As an example, in an energy analysis on 104 Austrian WWTPs, Haslinger et al. [41] found that for plant with design capacity lower than 100,000 PE the pretreatment impact for 12%, secondary treatment for 67% and sludge treatment for 15% of the total energy consumption, while for plants with design capacity higher than 100,000 PE the same plant sections the distribution of energy use relative to the total was respectively 11, 60 and 23%. Similar results are reported in other sources [42]. 23 Table 7. Weights of different KPIs to the overall energy consumption of a WWTP. Stage S1 S2 S3 S4 S5 Rapid Audit Value (wi) 0.119 - 0.535* 0.121 0.225 Decision Support Value (wi) 0.119 0.015 0.519 0.121 0.225 * In the RA, the function solid removal has been considered in stage 3 instead of stage 2 (as in DS) through COD removal, which like TS is a proxy of organic matter. As a result, the weight of stage 3 in RA is equal to the sum of the weight of stage 2 and stage 3 of the DS. If not all the KPIs are applicable, i.e. in the absence of one stage, weights should be normalised by the weights to sum unity such as described in Equation 8. 𝑤𝑛𝑜𝑟𝑚,𝑖= 𝑤𝑖 ∑𝑤𝑖 𝑘 1 (8) Where 𝑘 is the number of applicable KPIs. 3.2.4 Step 4: Aggregation Finally, aggregation consists in the combination of the weighted KPIs at either the stage or the whole plant level so that the corresponding WTEI can be computed and results compared based on a ranking. Aggregate the EPI into a single WTEI through a weighted sum (Eq. 9). 𝑊𝑇𝐸𝐼= ∑𝑤𝑛𝑜𝑟𝑚,𝑖 𝑘 𝑖=1 𝐸𝑃𝐼𝑖 (9) 3.2.5 Step 5: Rank and label assignation Using the cumulative frequency distribution curve of WTEI values allows the use of the percentile as an indicator of the energy efficiency performance. At this point, labelling is equivalent to assigning percentile intervals (bands) to energy classes. The scale is defined by fixing the transition values 24 between classes. The boundaries between labels (Table 7) have been decided according to the following criterion, common in EU efficiency labelling standards [43]: the median performance index is the upper boundary of class D. This labelling strategy allows good discrimination power at high efficiency, serving as an incentive for innovation. Table 7. Label definition according to the WTEI value, with A being the most energy efficient and G the least energy efficient. Label WTEI EPI1 EPI2 EPI3 EPI4 EPI5 A X<0.110 X<0.110 X<0.140 X<0.110 X<0.060 X<0.160 B 0.110≤X<0.220 0.110≤ X<0.220 0.140≤ X<0.280 0.110≤ X<0.220 0.060≤ X<0.120 0.160≤ X<0.320 C 0.220≤X<0.330 0.220≤ X<0.330 0.280≤ X<0.430 0.220≤ X<0.330 0.120≤ X<0.180 0.320≤ X<0.480 D 0.330≤X<0.440 0.330≤ X<0.440 0.430≤ X<0.560 0.330≤ X<0.440 0.180≤ EPI4<0.240 0.480≤ X<0.640 E 0.440≤X<0.550 0.440≤ X<0.550 0.560≤ X<0.700 0.440≤ X<0.550 0.240≤ X<0.300 0.640≤ X<0.800 F 0.550≤X<0.775 0.550≤ X<0.775 0.700≤ X<0.850 0.550≤ X<0.775 0.300≤ X<0.650 0.800≤ X<0.900 G X≥0.775 X≥0.775 X≥0.850 X≥0.775 X≥0.650 X≥0.900 3.3 Application of the ENERWATER methodology In this Section, the usefulness of the ENERWATER methodology is demonstrated step-by-step (Fig. 4) by using a real WWTP as an example, having a capacity of 35,800 Person Equivalent (PE) that removed N and P on top of COD, i.e. Type 2. The plant, which is further described in Section 4 of Supplementary Material, consumed a total of 3,575 kWh/d of energy (gross consumption), of which about 25% was due to the chemicals mainly used for P removal. Additionally, 1,409 kWh/d of the 25 plant electricity demand was balanced by biogas production, so its net energy consumption was 2,166 kWh/d. 32 organic matter, nitrogen, phosphorus etc.) do not vary significantly between WWTPs, hence restricting the application of these approaches in large geographical areas characterized by a wide heterogeneity in influent and effluent characteristics. As a result, there are not universal energy efficiency indicators that can be applied in every situation. Although the claim that “every plant is different” is shown to be correct, the methodology presented here represents a successful attempt to take into account this complexity by defining plant’s functions and corresponding energy efficiency indicators for each function. In this way plant energy performance is better represented and allows WWTPs saving energy while providing the desired level of wastewater treatment services; in other words, being more energy efficient. The methodology here presented allows for the first time the different stakeholders involved in the water sector to obtain and share standardized and comparable WWTP energy efficiency information, which has been previously identified as a major obstacle to reduced energy use at WWTPs. In particular by using this method engineers can test and compare energy saving strategies from different studies or plants, wastewater operators can properly evaluate the performance change after the implementation of any energy saving measure and decision-makers can employ a single energy label that uses universally known illustrations, to widely communicate information on the WWTP energy status. 5.2 Utility As far as the different stakeholders involved in the wastewater sector are concerned, it is likely that for the decision-making process easy and simple way to communicate energy efficiency level is necessary. A continuous exchange of experience at international level is in fact crucial to achieve the target of the Energy Efficiency Directive [7]. In doing so, countries may learn from each other’s experience and try to adopt best practices or at least avoid bad ones. The WTEI described in this paper represents a determined attempt to create a composite index, the WTEI, able to measure the multidimensional concept of energy efficiency at WWTPs. Composite indexes are in fact easier to interpret than a battery 33 of many separate indicators and facilitate communication among different stakeholders [38]. Having this object in mind, an energy label system has been developed taking into account that energy labelling is accepted and normalized at the present time in the private consumer sector and begins also to spread in the public sector. With the advent of Green Public Procurement (GPP), public administrations integrate environmental criteria at all stages of the purchasing process, encouraging the diffusion of sustainable technologies and the development of environmentally valid products, through research and choice of results and solutions that have the lowest possible impact on the environment throughout the entire life cycle [50]. Following the successful introduction of EU energy labelling for energy consuming devices and buildings, we argue that extending energy labelling to WWTPs would positively contribute to improving the exchange of information on energy saving actions and results between wastewater utilities and towards other stakeholders, thus supporting the concept of GPP. Any successful energy saving project must be based on a decision support framework able to identify sources of inefficiency and to assist plant operators in the decision-making process by suggesting energy saving actions. Nevertheless, tools limited to energy efficiency benchmarking cannot be considered diagnostic tools because they fail at prescribing any improvement strategy. The developed ENERWATER DS methodology is proposed to address this gap by intending to identify where inefficiencies come from in the plant. In fact, when diagnostic tools are reported in literature are in general too complex to be applied on a large scale due to the large amount of data and time required, as well as specific for some equipment. On the contrary, the RA ENERWATER methodology just requires parameters regularly measured in the plant. This quick assessment can facilitate the process of energy diagnosis, at least at the initial phase of inefficiency identification, by providing plant operators with case-based suggestions for energy efficiency. 5.3 Limitations 34 One limitation of the developed method is the availability of data. The 50 ENERWATER case studies were selected in order to cover the maximum range of the most widely used wastewater treatment techniques and reflecting the actual size distribution of European plants, whose majority are of medium-small size (i.e. less than 2,000 PE). Even if additional 48 WWTPs, whose data were retrieved from literature, were included in the final database, reaching a final dataset of 98 WWTPs, the number of observations is relatively small to be representative of all European WWTPs. Furthermore, for the sake of completeness and with the aim of designing a methodology that can be applied in the future as the complexity of WWTPs increases, the division into stages and the definition of KPIs has been done comprehensively, i.e. by defining additional stage 6 and 7 (respectively for return liquor and odour treatment) or by identifying KPIs for micropollutants [35]. However, not all the KPIs and stages can be, at the current state of development, combined into the WTEI. The lack of actual data on the contribution of each of these functions to the overall energy efficiency of the plants prevents from using them in the determination of WTEI. These indicators are kept, nonetheless, for future extensions of the DS ENERWATER methodology. Finally, influent composition and flowrate, together with other factors related to climate, location, etc., have a major impact on energy efficiency. There are techniques, e.g. based on regression analysis, which would allow estimating the fraction of inefficiency that can be traced back to the influent composition, and it would be possible to estimate inefficiency after controlling for the impact of the influent and other factors [25]. However, this is beyond the scope of the ENERWATER methodology. 5.4 Future outlook The study of the standardization landscape at European and at international level confirms the absence of specific normative documents in the framework of energy efficiency in wastewater treatment plants. Therefore, there is a good opportunity to fill this gap by raising a proposal based on the results of the presented work to the standardization organizations. To impulse dialogue towards the creation of a 35 specific European Standard, the corresponding standardization bodies were contacted (CEN/TC 165 at European level, CTN 149 at national level (Spanish)) so that the ENERWATER methodology could be the basis for a standardization document. As a result of a very favourable reception by CEN/TC 165, the ENERWATER methodology is being adapted to the European Technical Report format that will be submitted for consultation and voting by the CEN national members. If finally approved, this Technical Report could be the first step for a future European standard on energy efficiency in WWTPs. The European as well as national evaluation and monitoring process indicated by the Energy Efficiency Directive (Article 8) offers a window of opportunity for data collection purposes [7], which once being in a standardized form will favour the future design of policy instruments. These actions should also bring to European water industry a competitive advantage in new products development and a faster access to markets by facilitating evidence of energy reduction therefore fostering adoption on new technologies. 6. Conclusions This paper describes the first methodology specifically tailored to estimate energy efficiency at wastewater treatment plants. Starting from a clear definition of energy efficiency, the proposed methodology illustrates an innovative way to measure such energy efficiency by developing a tool for benchmarking and diagnosing the use of energy and formulating improvement actions based on previous analyses. The ENERWATER methodology was built up following a transparent procedure (public deliverables, stakeholder events, national and internal conference participations) that involved various stakeholders (universities, water utilities, standardisation bodies, SMEs and engineered product manufacturers), thus achieving a high-shared consensus in the industry. The main contributions of ENERWATER as a standard energy efficiency methodology for WWTPs are: i) accurate definition of WWTPs functions by identification of KPIs that reflect the operational 36 efficiency of each function, ii) ability to adapt to different plant layouts, iii) consideration of energy produced onsite; iv) good geographical coverage at European level, and v) easy communication by an aggregated indicator that reflect the complexity of a WWTP, the Water Treatment Energy Index. The case study illustrates the procedure to carry out an energy analysis and the usefulness of the proposed methodology for estimate the energy label of a WWTP. Additionally, the efficiency estimates obtained with the proposed methodology have been successfully validated with other techniques commonly employed in the literature, therefore suggesting a high level of robustness of the efficiency estimates produced by the ENERWATER methodology. Finally, it is interesting to remark that the proposed methodology can be easily applied by operators in existing WWTPs given that requires the measurement of common parameters generally measured in the plant, therefore it is expected that its application will facilitate the process of energy diagnosis, at least at the initial phase of inefficiency identification, by providing plant operators with case-based suggestions for energy efficiency. Moreover, we argue that extending energy labelling to WWTPs would positively contribute to improving the exchange of information on energy saving actions and results between wastewater utilities and towards other stakeholders, which is seen as crucial to achieve the target of the Energy Efficiency Directive. Acknowledgments We thank Benedetto Mirko d’Antoni and Diego Cingolani for their contribution to the development of the ENERWATER methodology. Stefano Longo, Miguel Mauricio-Iglesias and Almudena Hospido belong to the Galician Competitive Research Group (ED431C 2017/029) and the CRETUS strategic partnership (AGRUP2017/01), co-funded by FEDER (EU). Besides, they are supported by ‘ENERWATER’ Coordination Support Action (www.enerwater.eu) that has received founding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 649819. The authors would also like to thank the reviewers for their valuable comments. 37 References [1] Gu, Y., Li, Y., Li, X., Luo, P., Wang, H., Robinson, Z.P., Wang, X., Wu, J., and Li, F. (2017). The feasibility and challenges of energy self-sufficient wastewater treatment plants. Appl. Energy 204, 1463-1475. https://doi.org/10.1016/j.apenergy.2017.02.069. [2] Wang, H., Yang, Y., Keller, A.A., Li, X., Feng, S., Dong, Y., and Li, F. (2016). Comparative analysis of energy intensity and carbon emissions in wastewater treatment in USA, Germany, China and South Africa. Appl. Energy 184, 873-881. https://doi.org/10.1016/j.apenergy.2016.07.061. [3] Cohen-Shacham, E., Walters, G., Janzen, C., and Maginnis, S. (2016). Nature-based solutions to address global societal challenges. Gland: IUCN. Retrieved from www.portals.iucn.org/library/sites/library/files/documents/2016-036.pdf [4] Wakeel, M., Chen, B., Hayat, T., Alsaedi, A., and Ahmad, B. (2016). Energy consumption for water use cycles in different countries: A review. Appl. Energy 178, 868-885. https://doi.org/10.1016/j.apenergy.2016.06.114. [5] Krampe, J. (2013). Energy benchmarking of south Australian WWTPs. Water Sci. Technol. 67(9), 2059-2066. https://doi.org/10.2166/wst.2013.090. [6] EC (2012). Directive 2012/27/EU of the European Parliament and of the Council of 25 October 2012 on energy efficiency, amending Directives 2009/125/EC and 2010/30/EU and repealing Directives 2004/8/EC and 2006/32/EC. Brussels: European Council. Retrieved from www.eurlex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32012L0027&from=en. 38 [7] Nabitz, L., and Hirzel, S. (In Press). Transposing the requirements of the energy efficiency directive on mandatory energy audits for large companies: A Policy ‐ Cycle ‐ based review of the national implementation in the EU-28 member states. Energy Policy. https://doi.org/10.1016/j.enpol.2017.12.016. [8] EC (2000). Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for Community action in the field of water policy. Brussels: European Council. Retrieved from www.eur-lex.europa.eu/resource.html?uri=cellar:5c835afb-2ec6-4577-bdf8756d3d694eeb.0004.02/DOC_1&format=PDF. [9] Metcalf, L. and Eddy, H. P. (2003). Wastewater Engineering: Treatment and Reuse. New York: McGraw Hill. [10] DWA (2015). Standard DWA-A 216E: Energy check and energy analysis – instruments to optimise the energy usage of wastewater systems. Hennef: German Association for Water, Wastewater and Waste (DWA). [11] Longo, S., d’Antoni, B. M., Bongards, M., Chaparro, A., Cronrath, A., Fatone, F., Lema, J. M, Mauricio-Iglesias, M., Soares, A., and Hospido, A. (2016). Monitoring and diagnosis of energy consumption in wastewater treatment plants. A state of the art and proposals for improvement. Appl. Energy 179, 1251-1268. https://doi.org/10.1016/j.apenergy.2016.07.043. [12] Torregrossa, D., Hernández-Sancho, F., Hansen, J., Cornelissen, A., Popov, T., and Schutz, G. (2017). Energy saving in wastewater treatment plants: A plant-generic cooperative decision support system. J. Clean. Prod. 167, 601-609. https://doi.org/10.1016/j.jclepro.2017.08.181. [13] Torregrossa, D., Hansen, J., Hernández-Sancho, F., Cornelissen, A., Schutz, G., and Leopold, U. (2017). A data-driven methodology to support pump performance analysis and energy efficiency optimization in waste water treatment plants. Appl. Energy 208, 1430-1440. https://doi.org/10.1016/j.apenergy.2017.09.012. 39 [14] Pérez-Lombard, L., Ortiz, J., González, R., and Maestre, I. R. (2009). A review of benchmarking, rating and labelling concepts within the framework of building energy certification schemes. Energy Build. 41(3), 272-278. https://doi.org/10.1016/j.enbuild.2008.10.004. [15] May, G., Barletta, I., Stahl, B., and Taisch, M. (2015). Energy management in production: A novel method to develop key performance indicators for improving energy efficiency. Appl. Energy 149, 46-61. https://doi.org/10.1016/j.apenergy.2015.03.065. [16] Li, M., and Tao, W. (2017b). Review of methodologies and polices for evaluation of energy efficiency in high energy-consuming industry. Appl. Energy 187, 203-215. https://doi.org/10.1016/j.apenergy.2016.11.039. [17] Ahn, J., Cho, S., and Chung, D.H. (2016). Development of a statistical analysis model to benchmark the energy use intensity of subway stations. Appl. Energy 179, 488-496. https://doi.org/10.1016/j.apenergy.2016.06.065. [18] Salvatori, S., Benedetti, M., Bonfà, F., Introna, V., and Ubertini, S. (2018). Inter-sectorial benchmarking of compressed air generation energy performance: Methodology based on real data gathering in large and energy-intensive industrial firms. Appl. Energy 217, 266-280. https://doi.org/10.1016/j.apenergy.2018.02.139. [19] Panepinto, D., Fiore, S., Zappone, M., Genon, G., and Meucci, L. (2016). Evaluation of the energy efficiency of a large wastewater treatment plant in Italy. Appl. Energy 161, 404-411. https://doi.org/10.1016/j.apenergy.2015.10.027. [20] Di Fraia, S., Massarotti, N., and Vanoli, L. (2018). A novel energy assessment of urban wastewater treatment plants. Energy Convers Manag. 163, 304-313. https://doi.org/10.1016/j.enconman.2018.02.058. 40 [21] Geng, Z., Dong, J., Han, Y., and Zhu, Q. (2017). Energy and environment efficiency analysis based on an improved environment DEA cross-model: Case study of complex chemical processes. Appl. Energy 205, 465476. https://doi.org/10.1016/j.apenergy.2017.07.132. [22] Fernández, D., Pozo, C., Folgado, R., Jiménez, L., and Guillén-Gosálbez, G. (2018). Productivity and energy efficiency assessment of existing industrial gases facilities via data envelopment analysis and the Malmquist index. Appl. Energy 212, 1563-1577. https://doi.org/10.1016/j.apenergy.2017.12.008. [23] Zhang, G., and Lin, B. (2018). Impact of structure on unified efficiency for Chinese service sector—A twostage analysis. Appl. Energy 231, 876-886. https://doi.org/10.1016/j.apenergy.2018.09.033. [24] Molinos-Senante, M., and Sala-Garrido, R. (2018). Evaluation of energy performance of drinking water treatment plants: Use of energy intensity and energy efficiency metrics. Appl. Energy 229, 1095-1102. https://doi.org/10.1016/j.apenergy.2018.08.102. [25] Longo, S., Hospido, A., Lema, J. M., and Mauricio-Iglesias, M. (2018). A systematic methodology for the robust quantification of energy efficiency at wastewater treatment plants featuring data envelopment analysis. Water Res. 141, 317-328. https://doi.org/10.1016/j.watres.2018.04.067. [26] Chung, W. (2011). Review of building energy-use performance benchmarking methodologies. Appl. Energy 88, 1470-1479. https://doi.org/10.1016/j.apenergy.2010.11.022. [27] Chitnis, M., and Sorrell, S. (2015). Living up to expectations: Estimating direct and indirect rebound effects for UK households. Energy Econ. 52, S116. https://doi.org/10.1016/j.eneco.2015.08.026. [28] Liao, N., and He, Y. (2018). Exploring the effects of influencing factors on energy efficiency in industrial sector using cluster analysis and panel regression model. Energy 158, 782-795. https://doi.org/10.1016/j.energy.2018.06.049. 41 [29] Broin, E.Ó, Nässén, J., and Johnsson, F. (2015). Energy efficiency policies for space heating in EU countries: A panel data analysis for the period 1990–2010. Appl. Energy 150, 211-223. https://doi.org/10.1016/j.apenergy.2015.03.063. [30] Filippini, M., and Greene, W. (2016). Persistent and transient productive inefficiency: a maximum simulated likelihood approach. J. Prod. Anal. 45, 187-196. https://doi.org/10.1007/s11123-015-0446-y. [31] Longo, S., Chitnis, M., Mauricio-Iglesias, M., and Hospido, A. (2018). Transient and persistent energy efficiency in the wastewater sector based on economic foundations. Manuscript submitted for publication. [32] Carlson, S. W., and Walburger, A. (2007). Energy index development for benchmarking water and wastewater utilities. Denver: American Water Works Association (AWWA). Retrieved from www.waterrf.org/PublicReportLibrary/91201.pdf. [33] Revollar, S., Vega, P., Vilanova, R., and Francisco, M. (2017). Optimal Control of Wastewater Treatment Plants Using Economic-Oriented Model Predictive Dynamic Strategies. Appl. Sci.7(8), 813. https://doi.org/10.3390/app7080813. [34] Rodriguez-Garcia, G., Molinos-Senante, M., Hospido, A., Hernández-Sancho, F., Moreira, M. T., and Feijoo, G. (2011). Environmental and economic profile of six typologies of wastewater treatment plants. Water Res. 45(18), 5997-6010. https://doi.org/10.1016/j.watres.2011.08.053. [35] ENERWATER (2017). ENERWATER methodology. ENERWATER project, deliverable D3.4. Retrieved from www.enerwater.eu. [36] ENERWATER (2015). Study of published energy data. ENERWATER project, deliverable D2.1. Retrieved from www.enerwater.eu. [37] ENERWATER (2016). Benchmarking database. ENERWATER project, deliverable D2.5. Retrieved from http://www.enerwater.eu/.