Assessing the thermodynamic performance of Irish municipal wastewater treatment plants using exergy analysis: a potential benchmarking approach

Accepted Manuscript Benchmarking the thermodynamic performance of Irish municipal wastewater treatment plants using exergy analysis L. Fitzsimons, M. Horrigan, G. McNamara, E. Doherty, T. Phelan, B. Corcoran, Y. Delauré, Eoghan Clifford PII:

S0959-6526(16)30471-1

DOI:

10.1016/j.jclepro.2016.05.016

Reference:

JCLP 7198

To appear in:

Journal of Cleaner Production

Received Date: 15 September 2015 Revised Date:

4 May 2016

Accepted Date: 4 May 2016

Please cite this article as: Fitzsimons L, Horrigan M, McNamara G, Doherty E, Phelan T, Corcoran B, Delauré Y, Clifford E, Benchmarking the thermodynamic performance of Irish municipal wastewater treatment plants using exergy analysis, Journal of Cleaner Production (2016), doi: 10.1016/ j.jclepro.2016.05.016. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Benchmarking the thermodynamic performance of Irish municipal wastewater treatment plants using exergy analysis 1

L.Fitzsimons*,1 School of Mechanical and Manufacturing Engineering Dublin City University, Dublin, Ireland e−mail: [email protected]

SC

RI PT

M.Horrigan1, G.McNamara1, E. Doherty2, T.Phelan1, B.Corcoran1, Y.Delauré1 and Eoghan Clifford2 1 School of Mechanical & Manufacturing Engineering Dublin City University, Dublin, Ireland 2 College of Civil Engineering and Informatics, Ryan Institute, National University of Ireland, Galway, Ireland

ABSTRACT

AC C

EP

TE D

M AN U

Wastewater treatment is a resource intensive process utilising several inputs such as energy, chemicals and water to produce an effluent that meets designated environmental standards. It is predicted that increasingly stringent discharge requirements will include better treatment of nutrients and the need to treat emerging pollutants, i.e. pollutants not commonly monitored in the environment currently, but which have the potential to cause environmental damage. One expected outcome of increased effluent standards is higher energy consumption. Coupled with this, fluctuating energy costs and the need to mitigate fossil fuel based carbon emissions have led to a greater focus on the energy efficient operation of wastewater treatment plants. Exergy Analysis has been identified as a useful tool to assess the resource efficiency of thermal and other energy systems. The exergy approach provides a rational basis for process optimisation, where, in theory, the processes with the greatest exergy destruction represent the greatest energy efficiency opportunities. In this research, exergy analysis has been used to assess and compare the energy/resource efficiency of two Irish wastewater treatment plants. One important objective is to investigate the appropriateness of exergy analysis as a potential benchmarking approach for wastewater treatment plants. Although the two wastewater treatment plants are of similar scale, and use similar technologies, the results of the exergy analysis were significantly different. For example, the secondary treatment exergy destruction differed by as much as 67% between the two wastewater treatment plants. Furthermore, the coupled pre-treatment and secondary treatment exergetic efficiencies of both plants differed significantly: one wastewater treatment plant had an exergetic efficiency of 27.5% in comparison with 40.2% for the second plant. One important mitigating factor was identified: the difference between their incoming wastewater concentrations. These results indicate that metrics such as exergy destruction when combined with consideration of water quality can be a useful tool in the comparing the overall performance of wastewater treatment plants. KEYWORDS

Exergy analysis; wastewater treatment; resource efficiency; energy efficiency; benchmarking. 1. INTRODUCTION Wastewater treatment is a resource intensive process, with three main resources being identified as those of greatest concern: energy, chemicals and water. According to the United States Environmental Protection Agency (United States EPA, 2006), wastewater treatment accounts for approximately 1% of the world’s total energy consumption and 3% of the

ACCEPTED MANUSCRIPT

RI PT

electrical load in the United States. This 1% current estimation is mirrored in Europe, where energy consumption is expected to increase significantly due to population growth and increasing environmental standards (Olsson, 2012). It has been estimated that increases in wastewater treatment plant (WWTP) energy efficiency of over 20% could occur in the US by 2020 (United States EPA, ), whereas in Europe ‘conservative’ increases of 60 to 100% over the next 15 years are expected to meet the new EU directives (Olsson, 2012). Some treatment companies in the UK have reported increases in electricity usage of 60% since 1990 (Olsson, 2012). Coupling these trends and predictions with recent energy cost increases means that energy is, and increasingly will be, one of the major operational costs that many WWTPs face.

M AN U

SC

Driven by environmental regulations, the focus of WWTPs has traditionally been the quality of the effluent and not necessarily energy or resource efficiency. Regulation and penalties incentivise the meeting of environmental effluent standards, however, to date, there are no such analogous penalties or incentives to expedite the focus on resource efficiency. It is important to recognise that resource utilisation has significant environmental consequences also, and it is necessary to consider WWTP performance holistically. Note that, even with the effluent quality penalties/incentives in place, many Irish treatment WWTPs are not meeting minimum EU standards (Shannon et al., 2014), and this is replicated across the European Union (European Commission, 2013).

AC C

EP

TE D

Exergy is a thermodynamic property that combines the first and second laws of thermodynamics. A system can do work due to the potential difference in thermodynamic state that may exist between the system and its environment (or dead state). This potential difference can exist because of differences in temperature, pressure and chemical composition inter alia. Exergy analysis is an analytical method that has been widely used to assess and quantify the thermodynamic performance of energy and water systems, for example, thermal power plants and desalination plants. Exergy analysis typically involves four stages: (1) defining the system under consideration, generally modelled as a control volume for open systems; (2) quantifying the exergy of process inputs and outputs; (3) conducting an exergy balance across various processes or systems of interest to determine the exergy destruction and exergy losses that take place; (4) establishing a hierarchy of process exergy destruction and losses in systems and calculating the exergetic efficiency of the process or system. Methodologies for the assessment of WWTPs have been developed previously by several researchers (Tai et al., 1986) and have been used to estimate the consumption of physical resources in a WWTP in Sweden (Hellström, 1997) and to assess and optimise the thermodynamic efficiency of a WWTP in Iran (Khosravi et al., 2013). The research presented here forms part of a larger study to assess, benchmark and improve the resource efficiency of Irish WWTPs. The research hypothesis is defined as follows: exergy analysis is a useful analytical approach to assess, compare and potentially benchmark the performance of WWTPs. Specifically, the objectives of this paper are: (1) to review previous approaches to WWTP exergy analyses; (2) to undertake exergy analyses of two Irish WWTPs using available simultaneous water quality and resource data; (3) based on the results of the exergy analysis, to assess and compare WWTP thermodynamic performance; (4) to investigate and assess the appropriateness of exergy analysis as a potential benchmarking approach; (5) based on the results of the case study, to assess the implications for policy making decisions. 2. WASTEWATER EXERGY STUDIES

ACCEPTED MANUSCRIPT Initial work by Tai et al. (Tai et al., 1986) to assess the chemical exergy values of organic matter in waste water related the chemical exergy of organic matter to common waste water indices Total Oxygen Demand (TOD) and Total Organic Carbon (TOC) using Eq.s (1) and (2) below, eCh ( J / l ) = 13.6 ( kJ / g ) × TOD ( mg / l ) (1) (2)

RI PT

eCh ( J / l ) = 45 ( kJ / g ) × TOC ( mg / l ) Ch

TE D

M AN U

SC

where e is the chemical exergy of the wastewater stream. This was achieved by developing correlations between the calculated chemical exergy of 138 organic compounds (consisting of carbon, hydrogen, oxygen, chlorine, nitrogen and sulphur) and TOD and TOC. Tai stated that organic matter parameters Biochemical Oxygen Demand (BOD) and Chemical Oxygen Demand (COD) could also be used as approximate measures of effective energy because TOD indirectly represented the magnitude of utilisable energy from wastewater. A clear link exists between theoretical TOD and measured COD, where, for wastewaters COD is deemed approximately equivalent to theoretical TOD (Roberts Alley, 2007, Tai et al., 1986). No clear, general link, however, has been established between BOD and theoretical TOD. The actual oxygen demand of any organic compound is its biodegradability. Consequently, Hellström (Hellström, 1997) proposed the use of 7 day BOD (BOD7) in place of theoretical TOD because BOD7 was a better representation of the “amount of easily biodegradable organic matter”. Due to the extant lack of a clear link between BOD and theoretical TOD, COD will be used to estimate the chemical exergy of organic matter in waste water in this research. The chemical exergy of sludge, wastewater and effluent in this research will be calculated using Eq. 3 below. The use of this correlation-based equation to calculate the chemical exergy of wastewater organic matter is in line with the approaches of Tai (Tai et al., 1986), Hellström (Hellström, 1997) and Khosravi (Khosravi et al., 2013), although the latter use an amended equation. eCh ( J / l ) = 13.6 ( kJ / g ) × COD ( mg / l ) (3)

AC C

EP

Khosravi (Khosravi et al., 2013) extended the correlation analyses of Tai to include long chain compounds including nitrogen, which the authors stated were more typical of urban and industrial wastewater compounds. Their analysis resulted in the following correlation, see Eq. 4, which has been amended to include COD in place of the originally presented theoretical TOD. eCh ( J / l ) = 13.7 ( kJ / g ) × COD ( mg / l ) − 116 (4) The approach taken by Tai to calculate the chemical exergy of elements, which were subsequently used to develop the correlation in Eq. 3, was different to the global reference environment proposed by Szargut (Szargut et al., 1988, Szargut et al., 2005, Szargut, 1989). This is an important point because Szargut’s chemical exergy values are used by Khosravi to develop the correlation of Eq. 4. Szargut proposes a global reference environment with a defined, specific reference substance for each element, based on his proposed criterion. Specifically, Tai proposed an alternative reference environment, based on Yoshida et al. (Yoshida et al., 1984), which is summarised in Figure 1.

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 1: Approach to calculate the chemical exergy of organic compounds and to develop correlations

AC C

EP

TE D

The chemical exergy values of several elements and molecules, determined using each of the approaches, are presented in Table 1. Several elements and molecules have different chemical exergy values, notably water, chlorine and sulphur. This would imply that if any relevant molecules under consideration did contain these elements, then these correlations would change, not only because of the inclusion of long-chain compounds, but also because of the different chemical exergy values. The molecules used by Tai to develop the correlations do contain sulphur and chlorine. However, for the organic molecules considered by Khosravi to develop the correlations, both the long chain and short chain molecules, this is not the case. Therefore, it is only the inclusion of the additional long-chain compounds that leads to any differences between the two correlations (Eq. 3 and Eq. 4). One potential issue with Eq. 4 is that the chemical exergy values will be negative for COD values less than ~8.46 mg/l, which admittedly is less than the typical WWTP discharge limits of interest in this study. Another point of interest is that, in the case study presented by Khosravi, the nutrients, nitrogen and phosphorous, are calculated as separate inputs to the chemical exergy of COD. However, nitrogen has also been included in the calculation of the long-chain molecules to determine the chemical exergy correlations. This perhaps may lead to double counting. Typically, Total Nitrogen (TN) values consist of organic nitrogen (e.g. amino acids and proteins) that is readily converted to ammonium by microorganisms, ammonia, ammonium, nitrite and nitrate (Metcalf and Eddy, 2002). Table 1: Comparison of standard chemical exergy values (Szargut, 2005, Tai et al., 1986)

Element/molecule

H2 0 Cl2 S H2 C N2

Standard chemical exergy (kJ/mol) Szargut Tai et al. 9.5 123.6 609.6 236.1 410.3 0.7

0 46.9 602.8 235.2 410.7 0.72

ACCEPTED MANUSCRIPT

SC

RI PT

The impact of the choice of approach on the chemical exergy values for the organic matter was briefly assessed before undertaking the WWTP exergy analyses, see Table 2. It was found to be relatively insignificant at the higher incoming COD values (the average influent COD concentration at one of the plants under consideration in this research (WWTP E) was 426 mg/l over the testing period). However, the percentage differences between the two approaches increase for lower values of COD. For example, at the COD discharge requirements for both WWTPs (125 mg/l) considered, there is a 6.5% difference between the two chemical exergy values. One of the other WWTPs in the larger study had discharge requirements of 50 mg/l COD, which would result in a 19.5% difference between the two approaches. It is interesting that the reported inclusion of nitrogen by Khosravi to develop the chemical exergy correlation, discussed previously, results in lower chemical exergy values at the concentrations under consideration. Essentially, regarding the choice of approach, what needs to be determined is the nature of the wastewater at each WWTP and whether it comprises long-chain or short-chain molecules. As this is beyond the scope of the present paper, and based on the preceding discussion, the Tai correlation is used.

Chemical exergy (J/l) Khosravi et al. Tai et al. % diff. 569 680 -19.5 1597 1700 -6.5 1939 2040 -5.2 3309 3400 -2.8 5364 5440 -1.4 6049 6120 -1.2

TE D

COD (mg/l) 50 125 150 250 400 450

M AN U

Table 2: Comparison of chemical exergy using approaches of Tai and Khosravi (Khosravi et al., 2013, Tai et al., 1986)

AC C

EP

Regarding the exergy values for WWTP nutrients, Hellström (Hellström, 2003, Hellström, 1997) made the assumptions that all nitrogen within the WWTP exists as the ammonium ion and all phosphorus exists as ‘phosphate’ ( HPO42 − ) . The orthophosphate HPO42 − defined by Hellström is generally termed the hydrogen phosphate ion. These assumptions of nutrient state were also adopted by Khosravi. However, the chemical exergy values in both papers do not correspond. Hellström presents values for the ammonium ion (322.1 kJ/mol) and the hydrogen phosphate ion (134.1 kJ/mol), whereas the respective Khosravi values are 393.14 kJ/mol and -103.67 kJ/mol. The Hellström values, however, are not technically the chemical exergy values of the hydrogen phosphate and ammonium ions, but rather the standard chemical exergy values of dissociated electrolytes in aqueous solutions (i.e. one molal solutions of NH4OH and H3PO4 at standard temperature and pressure). The chemical exergy of ionic species or other compounds can be calculated using the following approach (Szargut et al., 1988), e Ch = ∆ F G ° + ∑ veiCh (5) i

where eCh is the standard molar chemical exergy of the compound or ion under consideration, ∆ F G ° is the standard Gibbs energy of formation of the ion or compound in question, v is the

ACCEPTED MANUSCRIPT stoichiometric coefficient of each element in the compound and ei is the standard molar chemical exergy of each element in the compound; the subscript i refers to each of the elements under consideration. The chemical exergy of the ammonium ion can be calculated as follows in Eq. 6, 1 Ch eNH + eNCh2 + 2eHCh2 (6) + = ∆FG NH 4+ 4 2

RI PT

substituting the relevant values in Eq. 7, where the Gibbs energy of formation value is taken from Wagman et al. (Wagman et al., 1982) and the chemical exergy values from Szargut (Szargut, 2005), the standard molar chemical exergy of the ammonium ion is 393.2 kJ/mol. 1 Ch eNH (0.69) + 2(236.1) = 393.2 kJ/mol (7) + = −79.31 + 4 2

(8)

M AN U

SC

The chemical exergy of the hydrogen phosphate ion is shown in Eq. 8. 1 Ch eHPO (236.1) + 861.4 + 2(3.97) = −101.76 kJ/mol 2− = −1089.15 + 4 2

These values are similar to the values reported in Khosravi, i.e. a value of 393.14 kJ/mol for the ammonium ion and -103.67 kJ/mol for the hydrogen phosphate ion. However, anions and cations exist as electrolyte pairs, and it is unusual to consider them in isolation. On this basis, the Szargut values of 134.1 kJ/mol for the hydrogen phosphate ion ( H 3 PO4 → 2 H + + HPO42− ) and 322.1 kJ/mol for the ammonium ion ( NH 3 → NH 4+ + OH − ) are chosen for this research.

AC C

EP

TE D

Martínez and Uche (Martínez and Uche, 2010) studied the chemical exergy of a water flow, where the main objective was to use exergy as an indicator of the depletion and degradation of natural resources. With a particular focus on the chemical exergy of organic matter, they compared the values obtained using the method proposed by Tai with other approaches to determine typical organic molecules to represent the chemical exergy of river water. An important part of the work involved the selection of an appropriate reference environment to quantify the degradation of a water body. The reference environment was specifically selected so that the exergy value indicated relative water purity; the authors studied a number of hypothetical reference environments to assess the impact of each of the reference environments on the corresponding exergy values. There was another important consideration in the determination of the appropriate reference environment, and that was whether or not to include organic matter in the reference environment. According to the authors’ proposed approach, the inclusion or not of a species determines whether the chemical exergy of a species in the water body is assessed on the basis of differences in concentration or on the basis of intrinsic chemical exergy, respectively. The reference environment deemed most appropriate was seawater, where the organic matter was excluded from the reference environment. The proposed exergy approach of Martínez is relevant to WWTPs because it specifically considers the impact of the water quality of a water body (e.g. WWTP effluent) as it mixes with its discharge environment, for example, rivers or the sea. However, this approach to calculate the chemical exergy of the water body requires consideration because it appears to be a splicing of the global reference environment developed by Szargut and the concentration differences typically adopted by the desalination exergy practitioners (Fitzsimons et al., 2015,

ACCEPTED MANUSCRIPT

RI PT

Sharqawy et al., 2011, Uche, 2000). Generally, in the latter research area, the chemical exergy at various process stages is solely based on the difference in concentration between the defined dead state and the process stage under consideration at dead state temperature and pressure. The approach of Martínez may lead to an anomaly or contradiction because the authors use the Szargut global reference environment on the one hand as the basis on which to calculate the chemical exergy of the organic matter and nutrients in the water body, but on the other hand use a local reference environment to assess the concentration differences for other species. The two approaches (local and global) were compared in previous work to assess a semiconductor ultra-pure water plant (Fitzsimons, 2011). The Szargut approach (Szargut et al., 1988) to calculate the chemical exergy of non-ideal mixtures or solutions is given in Eq. 9, eCh = ∑ xi eiCh + RT0 ∑ xi ln ai i

(9)

SC

i

EP

TE D

M AN U

where eCh , the standard molar chemical exergy, is given by the sum of the chemical exergies of each of the solution species and the entropy of mixing term. In Eq. 9, x is the mole fraction of the various species i , a is the activity of the various species i , R is the universal gas constant and T0 is the dead state temperature, in this case standard temperature. Equation 9 is an elegant expression in that the molar chemical exergy is pre-defined with respect to the global dead state. Without chemical changes, the chemical exergy differences across processes exist because of differences in the activities of the various species, in much the same way as the concentration differences in desalination exergy analyses. Note that Eq. 9 differs from the equation presented in Martínez in several ways. One difference is the second term on the right of the equation sign, where, rather than being pre-defined with respect to the global dead state, Martínez considers the concentration exergy of certain species relative to their defined reference environment. There is another difference, i.e. the chemical exergy term in Martínez is the Szargut equation to calculate the standard chemical exergy of compounds, whereas the first term in Eq. 9 is the sum of standard chemical exergy values of the elements in a solution. (Note that the standard chemical exergy of electrolytes and other typical solutes is calculated relative to the standard state for solutions, i.e. a one molal solution at standard temperature and pressure.)

AC C

It is evident that there are notable and interesting differences between the approaches proposed by various researchers that warrant further investigation. The goal of this paper is to use the approaches developed by Tai et al. and Hellström to undertake exergy analyses of two Irish WWTPs and to investigate the appropriateness of exergy analysis as a WWTP benchmarking tool. The key reason for using the Tai et al. and Hellström approaches is that there are a number of issues that have been raised in relation to the approach of Khosravi et al., particularly regarding the chemical exergy values of the electrolytes and the resulting negative values at low COD concentrations.

3. WWTP CHARACTERISTICS, DESCRIPTION AND TESTING This research forms part of a larger Irish EPA funded project to assess the holistic performance of Irish WWTPs from a number of synergistic perspectives: benchmarking, process control and optimisation, Life-Cycle Assessment (LCA) and Exergy Analysis. A

ACCEPTED MANUSCRIPT number of representative Irish WWTPs were chosen for the study, primarily based on their scale, discharge location and treatment technologies, see Table 3. Table 3: WWTP descriptions Agglomeration Served (PE)1

Receiving water body type

A B C D E F G H I J

186,000 25,000 24,834 18,517 12,000 12,000 5,000 820 750 600

79,133 18,659 22,440 25,633 12,284 9,036 2,500 590 422 1,024

Seawater Freshwater Freshwater Freshwater Freshwater Freshwater Freshwater Freshwater Freshwater Freshwater

Level of treatment (P),(S),(T)2 P,S P,S P,S,T P,S P,S P,S P,S P,S P,S P,S

Type of secondary treatment3 AS AS AS+P AS AS+P AS+P AS+P AS+P PFBR AS+P

RI PT

Design Capacity (PE)

SC

WWTP

M AN U

Two WWTPs were selected for the exergy analyses, WWTP E and WWTP F in Table 3. The reason for choosing these WWTPs is that they provide a ready comparison; they have similar technologies (conventional Activated Sludge (AS)), are of similar scale, have the same level of treatment and have similar discharge licence requirements, in terms of COD. The discharge requirements are shown in Table 4. Table 4: Discharge requirements

WWTP F

6-9

6-9

TE D

PH

WWTP E

CBOD

25mg/l

20mg/l

COD

125mg/l

125mg/l

SUSPENDED SOLIDS

35mg/l

30mg/l

-

20mg/l

2 mg/l

1 mg/l

5mg/l

-

1 mg/l

-

EP

TOTAL NITROGEN (as N) TOTAL PHOSPHORUS (as P) AMMONIA (as N)

AC C

ORTHOPHOSPHATE (as P)

Plant E comprises of screening, grit removal, three aeration tanks (diffused aeration system), two clarifiers, phosphorus removal, sludge thickening and sludge dewatering. Storm water storage tanks and a picket fence thickener are also included as part of the wastewater treatment works. The clarified effluent is discharged to a river. Plant F inlet works comprises mechanical and manual screens together with a compaction unit, overflow unit and grit traps. The influent is then passed to the anoxic tanks where it is mixed with return activated sludge (RAS). The effluent from each tank is spilt between the two aeration basins. The secondary treatment process is a single stage anoxic zone extended aeration process followed by clarification. The clarified effluent is discharged to a river. 1

Annual Environmental Report data. Agglomeration, as defined in the Waste Water Discharge (Authorisation) Regulations, means an area where the population or economic activities or both are sufficiently concentrated for a waste water works to have been put in place 2 P=Primary, S=Secondary, T=Tertiary 3 AS = Activated Sludge, PFBR = Pump Flow Biofilm Reactor, +P = with phosphorus removal

ACCEPTED MANUSCRIPT

RI PT

A detailed energy and water quality audit of both WWTPs was undertaken. Influent and effluent samples were taken at a maximum of 8 hour intervals in the case of grab samples or as daily composite samples, where each portion of the sample was collected at 4 hour intervals. Energy data and power quality data were gathered at intensive frequencies. Energy auditing was performed using the Fluke 435 Series II power quality analyser (PQA), which is a high-specification energy analyser. The PQA was supplemented with three Amprobe PQ 55A energy analysers. These devices are mid-range cost and specification and were capable of recording all basic variables. Finally, smaller WWTP equipment was metered using eight Iso-Tech IPM2005 meters. Table 5 outlines the basic specifications of each metering device in more detail. Table 5: Basic specifications for power/energy monitors utilised in WWTP audits Capability

Logger

Mains

Single and 3 phase

SD Card (8 GB)

Mains

Single and 3 phase

20000 records

Battery

Single and balanced 3 phase

8000 records

Sampling Freq. (Hz)

Harmonics (up to)

SC

Fluke 453 Series II Amprobe PQ 55A Iso-Tech IPM 2005

Power

Coms

1.3e-4 – 4

50th

USB

8.0e-3 – 0.2

31st

RS-232

1.6e-3 – 1

n/a

USB optical

M AN U

Monitor

TE D

Daily flow data were collected from the SCADA systems or daily logs. Water quality testing was carried out according to standard methods. Ammonium-nitrogen (NH4-N), total oxidised nitrogen (TON), nitrite-nitrogen (NO2-N), and phosphate-phosphorus (PO4-P) concentrations were determined using a Thermo Clinical Labsystems, Konelab 20 Nutrient Analyser (Fisher Scientific, Waltham, Massachusetts, United States). Suspended solids (SS) were measured in accordance with standard methods. Total Phosphorous (TP), TN, TOC and total inorganic carbon (TIC) were analysed using a BioTector TOC TN TP Analyser (BioTector Analytical Systems Limited, Cork, Ireland) in accordance with standard methods APHA 2005 (APHA, AWWA and WEF, 2005, 2005). BOD and COD were also measured in accordance with standard methods (APHA, AWWA and WEF, 2005, 2005). The measured energy and water quality data are presented in the Appendix, Table 1 and Table 2.

AC C

EP

4. EXERGY CALCULATIONS Whereas Martínez (Martínez and Uche, 2010) considers exergy as an indicator of environmental degradation, exergy analysis is primarily used to characterise, assess and optimise engineering system performance. In this research, exergy analyses were undertaken for the pre-treatment processes (screening, grit removal), the secondary treatment processes (aeration and secondary clarification), and an extended boundary analysis incorporating pretreatment and secondary treatment. The main reasons for selecting these specific processes and process configuration are: (1) the main energy driver in WWTPs is the aeration energy requirements; (2) data limitations at several of the audited plants, which are discussed later. In terms of engineering processes, what is of most interest is the exergetic efficiency of the WWTP, and the exergy destruction and exergy losses across processes and systems. It is important to distinguish exergy destruction and exergy losses in this context. Exergy destruction is calculated by performing an exergy balance across the various WWTP processes: exergy in minus exergy out equals exergy destruction; exergy losses are exergy streams that are outputs of WWTP processes, but are not utilised or harnessed in the WWTP to perform useful work. The rational exergetic efficiency (REE) definition used in this research is that of Kotas (Kotas, 1995): in general, this is defined as the exergy of the desired

ACCEPTED MANUSCRIPT output divided by the exergy of the required inputs. The relevant flows and overall WWTP energy data for the two WWTPs are presented in the appendix. All exergy flows including energy, the chemical exergy due to COD, and the chemical exergy due to the nutrients were converted to MJ/day to facilitate ready comparisons.

RI PT

One very important issue that became apparent during this research was the lack of calibrated monitoring equipment available at several of the WWTPs assessed in this study: few WWTPs had any level of energy monitoring; flow data was very often not available or unreliable; in some cases access to electrical panels to install energy meters on process equipment was not possible. Lack of data is a significant barrier to the effective management and optimisation of wastewater treatment WWTPs. It is also a barrier to conducting exergy analyses, where accurate sludge and energy data, emissions to air and other relevant information is desirable, but was often unavailable, resulting in a number of unavoidable estimations and assumptions.

AC C

EP

TE D

M AN U

SC

Calculation assumptions: • The processes are isothermal and isobaric; therefore, thermomechanical exergy is negligible. • Steady state is assumed. • Heavy metals and chemicals input data were not available and have been omitted form the analysis. • Average, measured influent and effluent COD data (mg/l) obtained during the testing period were used in the calculations where available. Inter-process COD data were not available but were estimated using values in the literature. • Average, measured energy values were used (kWh/day) for most processes; RAS pumping energy data were available at WWTP F but not at WWTP E; grit blower energy was available at WWTP E but not at WWTP F. Because of the similar flows and WWTP technologies, the energy consumption of the missing equipment was assumed to be equivalent for the two WWTPs. • The chemical exergy of the nutrients (kJ/mol) were calculated using the values in Szargut and were subsequently converted to MJ/day. Dimensional accuracy was ensured by dividing the chemical exergy by the molar mass and then multiplying the result by the product of the COD concentration and the daily flow rate; these values were finally converted to MJ/day. Molar mass values of NH4OH (35 g/mol) and H3PO4 (98 g/mol) were used in the conversions. • Emissions to air data were not available and were omitted from the analyses. • Inter-process nutrient data were not available, and therefore, nutrient data are not included in the pre-treatment and secondary treatment process exergy analyses. However, nutrient data were included in the extended boundary exergy analyses (discussed later). • Inter-process sludge data were not available. Average WWTP sludge output data was available from operator records; however, RAS and waste activated sludge (WAS) typical concentrations and flow rates were estimated using relevant values from the literature. • No data were available on the screenings or grit removal quantities; it was assumed that the volumetric flow rate entering the WWTP pre-treatment stage was equal to the flow rate after pre-treatment. COD reduction across processes relevant to this research is reported in Straub (Straub, 1989), i.e. a 5% to 10% COD reduction with fine screening and a 50% to 80% reduction following aeration and sedimentation. Based on measured values, COD % reduction values are 75.5%

ACCEPTED MANUSCRIPT and 73.5% for WWTP E and WWTP F respectively. Approximate midpoint values from Straub of 7.5% will be adopted initially for pre-treatment with the remaining COD % reduction allocated to the secondary treatment stage (aeration/secondary clarification).

SC

RI PT

In relation to sludge estimation, Metcalf and Eddy (Metcalf and Eddy, 2002) states that RAS flow rates of 50% to 75% of the average design flowrate are typical. According to the Irish EPA wastewater treatment manual (Irish EPA, 1997), RAS rates of up to 150% for smaller WWTPs are common. Regarding RAS concentrations, reported values vary widely: 4,000 mg/l to 12,000 mg/l (Metcalf and Eddy, 2002). Values of 4355 mg/l and 9,300 mg/l are reported elsewhere in the literature (Khosravi et al., 2013). In this research, midpoint values will be used initially to undertake the exergy analyses, that is, values of 7.5%, 8,000 mg/l and 90% for pre-treatment % COD reduction, sludge COD concentration and % RAS rates, respectively. This will be followed by an assessment of two alternative scenarios and a sensitivity analysis. The range of values is shown in Table 6. Table 6: Range of estimated values for COD

Estimated sludge COD concentration (mg/l)

Estimated % RAS rates (as % of influent flow)

5 7.5 10

4,000 8,000 12,000

60 90 120

AC C

EP

TE D

M AN U

Estimated % COD reduction following pretreatment

Figure 2: Secondary treatment process COD mass balance

For each of the scenarios, the unknown COD flow rate Q& 4 was calculated by COD mass balance (Figure 2) as follows in Eq. 10, Q&1C1 + Q& 3C3 = Q& 2C2 + Q& 4C4 (10) where Q& i are the volumetric flow rates and Ci are the COD concentrations of each of the process streams respectively. It is assumed that C4 = C3 = C5 . To calculate the exergy losses associated with the sludge output, a second mass balance equation (Eq. 11) was used to determine the WAS flow rate.

ACCEPTED MANUSCRIPT Q& 5C5 = Q& 4C4 − Q& 3C3

(11)

SC

RI PT

5. EXERGY ANALYSIS RESULTS AND DISCUSSION Exergy analyses of both WWTPs were carried out using the midpoint values in Table 6 initially, i.e. 7.5% COD reduction, 8,000 mg/l COD and 90% RAS rate. Note that the exergy analyses were undertaken for the pre-treatment processes, the secondary treatment processes, and for an extended boundary incorporating both processes (Figure 3). The exergy balance inputs and outputs for WWTP E and WWTP F are presented in Figure 4 and Figure 5 respectively. The exergy analysis results are compared in Table 7. The main reason for carrying out the extended boundary analysis was to include TN and TP; it was difficult to find reasonable estimates in the literature for TN and TP reduction across pre-treatment processes. The exergy flows used in the extended boundary analyses are the input and outputs shown at the extended boundary in Figure 3 in addition to the total energy inputs; the relevant values for WWTP E and WWTP F are given in Figure 4 and Figure 5 respectively.

AC C

EP

TE D

M AN U

At the pre-treatment stage the magnitude of exergy destruction in WWTP E is 31.3% greater than WWTP F. This is due to the difference in the incoming COD concentrations and the estimated changes in concentration that take place. (Note that the pre-treatment electrical energy inputs are assumed to be equivalent and that it is primarily the difference in COD concentrations that causes these differences.) The exergy due to the measured pre-treatment electrical energy requirements in both WWTPs is 180 MJ/day, whereas the exergy flows due to the influent COD concentrations are 10,709 MJ/day and 6,605.4 MJ/day respectively. Accordingly, the pre-treatment energy requirements are not a major factor in the pre-treatment exergy balance relative to the organic matter.

Figure 3: Pre-treatment, Secondary treatment and Extended boundaries for the exergy analyses

RI PT

ACCEPTED MANUSCRIPT

Pre-treatment Inputs: 3

3

SC

Wastewater (COD 426.1mg/l; Flow rate 1,848 m /day) Electricity: grit blowers (50.01 kWh/day) Outputs: Wastewater (COD 394.14 mg/l; Flow rate 1,848 m /day) Secondary treatment Inputs: 3

M AN U

Wastewater (COD 394.14 mg/l; Flow rate 1,848 m /day) Electricity: aeration blowers and sludge return pumps (1,366.9 kWh/day) 3

Return Activated Sludge (COD 8,000 mg/l; Flow rate 1,663.2 m /day Outputs: 3

Effluent (COD 104.53 mg/l; Flow rate 1,696 m /day)

3

Waste Activated Sludge (COD 8,000 mg/l; Flow rate 68.89 m /day) Extended boundary Inputs: 3

TE D

Wastewater (COD 426.1mg/l; Flow rate 1,848 m /day) Electricity: grit blower, aeration blowers and sludge return pumps (1,416.9 kWh/day) TN (71.46 mg/l) TP (7.66 mg/l) Outputs: 3

Effluent (COD 104.53 mg/l; Flow rate 1,696 m /day) 3

EP

Waste Activated Sludge (COD 8,000 mg/l; Flow rate 68.89 m /day) TN (50.06 mg/l) TP (0.98 mg/l)

AC C

Figure 4: WWTP E exergy analysis inputs and outputs

ACCEPTED MANUSCRIPT

RI PT

Pre-treatment Inputs: 3

Wastewater (COD 245.3mg/l; Flow rate 1,980 m /day) Electricity: grit blowers (50.01 kWh/day) Outputs: 3

3

SC

Wastewater (COD 226.9 mg/l; Flow rate 1,980 m /day) Secondary treatment Inputs:

Wastewater (COD 226.9 mg/l; Flow rate 1,980 m /day) Electricity: aeration blowers and sludge return pumps (450 kWh/day) 3

M AN U

Return Activated Sludge (COD 8,000 mg/l; Flow rate 1,782 m /day Outputs: 3

Effluent (COD 64.9 mg/l; Flow rate 1,944 m /day)

3

Waste Activated Sludge (COD 8,000 mg/l; Flow rate 40.39 m /day) Extended boundary Inputs: 3

TE D

Wastewater (COD 245.3 mg/l; Flow rate 1,980 m /day) Electricity: grit blower, aeration blowers and sludge return pumps (500.01 kWh/day) TN (29.6 mg/l) TP (3.63 mg/l) Outputs: 3

Effluent (COD 64.9 mg/l; Flow rate 1,944 m /day) 3

EP

Waste Activated Sludge (COD 8,000 mg/l; Flow rate 40.39 m /day) TN (16 mg/l) TP (0.85 mg/l) Figure 5: WWTP F exergy analysis inputs and outputs

AC C

Table 7: Comparison of exergy analysis results using midpoint values

WWTP

WWTP E WWTP F % difference

Exergy destruction (MJ/day) Pre-treatment Secondary Extended treatment boundary 983.2 675.4 31.3

4920.7 1619.9 67.1

6354.9 2556 59.8

Exergy losses (MJ/day)

Exergetic efficiency %

8278.5 4682.7 43.4

27.5 40.2 -46.2

At the secondary treatment stage, the magnitudes of exergy destruction are 4,920.7 MJ/day for WWTP E and 1,619.9 MJ/day for WWTP F, a percentage difference of 67%. Putting this in context, the COD exergy flows are 9,905.9 MJ/day for WWTP E and 6,110 MJ/day for WWTP F. Measured COD values were used for the effluent streams, resulting in exergy flows

ACCEPTED MANUSCRIPT

RI PT

of 2,411 MJ/day and 1,715.9 MJ/day respectively. The exergy flows due to the aeration and RAS pumping energy inputs are 4,920.7 MJ/day for WWTP E versus 1,619.9 MJ/day for WWTP F. Therefore, reviewing Table 7, it is evident that the exergy destruction that takes place in the secondary treatment process is equivalent to the exergy flows due to the electricity inputs. This is an interesting finding that will be considered in more detail later. The exergy losses, which are due to the WAS and the TN and TP outputs, again are notably different for both WWTPs: 8,278.5 MJ/day versus 4,682.7 MJ/day. The main reason for this is the difference in output COD concentrations.

SC

The hierarchy of exergy destruction that takes place across WWTP processes is important for focusing improvement efforts. For both WWTPs the majority of exergy destruction takes place at the secondary treatment stages, however, the proportion of total exergy destruction (i.e. the sum of pre-treatment and secondary treatment exergy destruction) attributable to the pre-treatment and secondary treatment processes is quite different. That is, the pre-treatment process in WWTP E is responsible for 16.7% of the total exergy destruction, whereas it is 29.4% for WWTP F.

TE D

M AN U

The exergetic efficiency of the extended boundary was calculated for both WWTPs. There are a number of exergetic efficiency definitions in the literature, which have been discussed in depth in previous publications (Cornelissen, 1997, Fitzsimons, 2011, Kotas, 1995). As mentioned previously, this work defines the REE as the exergy of the desired output divided by the exergy of the required inputs, which include the electrical exergy flows and any changes in exergy flows required to produce the desired output. Mathematically, for the WWTPs under consideration, this can be defined as, Exergyeffluent REE = (12) Exergyelectrical + Exergyinf luent + ExergyTN ,TP ,in − Exergysludge − ExergyTN ,TP ,out

The REE was found to be 27.5% for WWTP E and 40.2% for WWTP F. Again, this result is considered in detail later.

AC C

EP

The analyses were repeated for the low (5%; 4,000 mg/l; 60%) and high (10%; 12,000 mg/l; 120%) values in Table 6, and the results are shown in Tables 8 and 9 respectively. First, the pre-treatment processes for both WWTPs are considered. The magnitude of the exergy destruction rates increase and decrease in step with the low and high values, as do the percentage differences between the WWTPs. The secondary treatment results are interesting; the magnitudes of exergy destruction do not change across the high, low and mid-point scenarios and are again equivalent to the corresponding electrical exergy flows. Consequently, the exergetic efficiency values do not change. The pattern is repeated for the exergy losses. The REE increases at the low values and decreases with the higher values; this is expected because higher exergy destruction leads to lower exergetic efficiency. Table 8: Comparison of exergy analysis results using low values (5%; 4,000 mg/l; 60%)

WWTP WWTP E WWTP F % difference

Exergy destruction (MJ/day) Pre-treatment Secondary Extended treatment boundary 715.5 510.3 28.7

4920.7 1619.9 67.1

6087.2 2390.9 60.7

Exergy losses (MJ/day)

Exergetic efficiency %

8546.2 4847.8 43.3

28.4 41.8 -47.2

ACCEPTED MANUSCRIPT Table 9: Comparison of exergy analysis results using high values (10%; 12,000 mg/l; 120%)

WWTP E WWTP F % difference

1250.9 840.6 32.8

4920.7 1619.9 67.1

6622.7 2721.1 58.9

Exergy losses (MJ/day)

Exergetic efficiency %

8010.7 4517.5 43.6

26.7 38.7 -44.9

RI PT

WWTP

Exergy destruction (MJ/day) Pre-treatment Secondary Extended treatment boundary

SC

Other scenarios were trialled and particular consideration was given to the low influent concentrations in WWTP F and the impact this may have on the WWTP RAS strategy: for example, 60% RAS rate, 12,000 mg/l RAS concentration for WWTP E versus 120% RAS rate versus 4,000 mg/l RAS concentration for WWTP F. Based on a mid-point pre-treatment COD reduction, the results were identical to those presented in Table 7.

TE D

M AN U

One important finding that needs to be discussed further is the secondary treatment exergy analyses, where it was found that the process exergy destruction was equivalent to the electrical exergy inputs for each of the various scenarios considered during this study. This would suggest that the exergy flows due to the COD concentrations recirculate and transit the process without chemical exergy destruction. It is acknowledged that, due to the lack of measured inter-process sludge data, a number of assumptions were made with respect to the COD concentrations and RAS flow rates. However, the assumptions used in the mass balance are reasonable. The most likely cause is the COD concentration to chemical exergy conversion.

AC C

EP

The chemical exergy of each of the COD secondary streams is calculated using the Tai correlation, thus assuming that the chemical exergy of organic matter streams of equivalent COD concentrations are identical. This is the common approach in the literature. Indeed, with regard to this point, the exergy content of sludge/biosolids has been specifically discussed. The organic matter is assumed to exist as cellulose or cell tissue, depending on the ratio of nitrogen (Hellström, 1997, Khosravi et al., 2013). However, the sludge/biosolids were deemed suitable for simplification to organic matter because “the main part of the exergy in the biosolids is due to their organic matter” and the fractionation of the organic matter into cell tissue and cellulose does not lead to improvement in the results (Hellström, 1997). This simplification when used in the preceding exergy analyses effectively means that no change in chemical exergy takes place across the secondary treatment process. This warrants further investigation as it signifies that the issue could be a result of the model assumptions (i.e. COD and exergy are related purely by the concentration of COD without consideration of the changes in composition of the constituents that result in the COD measurements at various process stages). The REE of both WWTPs differs significantly. At the mid-point values the REE was found to be 27.5% for WWTP E and 40.2% for WWTP F, signifying that WWTP F has a better thermodynamic performance that WWTP E. The efficiency increases for both plants at the low values, and decreases at the high values (Table 8 and Table 9). Equation 12 requires consideration: the REE increases as the numerator increases (the exergy of the effluent), the denominator decreases (the electrical energy inputs decrease; the exergy

ACCEPTED MANUSCRIPT

RI PT

of the influent decreases; the change in exergy of the nutrients decreases; the exergy of the sludge increases), or a combination of the above. The exergy of the effluent is determined by the discharge requirements and the WWTP operator typically has no control over the influent. Therefore, this suggests that in order to run a WWTP that is thermodynamically efficient, one should try to just meet the discharge requirements while minimising the electrical energy inputs. Furthermore, this has implications for the future: more stringent discharge limits and the expected energy increases required to meet these limits will reduce the exergetic efficiency of WWTPs unless energy efficiency is given due consideration. This is an important point to consider for policy makers who may be unaware of the potential implications of stricter discharge limits, particularly in terms of additional energy and chemical requirements.

M AN U

SC

The differences in performance between the two WWTPs were notable across the three main scenarios on a number of levels: exergy destruction, exergy losses and exergetic efficiency. However, there is one very important mitigating factor, and that is the differences between the influent concentrations. WWTP E had an influent concentration of 426.1 mg/l COD versus 245.3 mg/l COD for WWTP F. Although the measured effluent COD was much lower for WWTP F (64.9 mg/l COD versus 104.5 mg/l COD), the electrical energy inputs for the extended boundary analysis were very different for both WWTPs (1416.9 kWh/day for WWTP E versus 500 kWh/day for WWTP F). The end result was very different exergetic efficiency profiles. As mentioned previously, the WWTP operator has little or no control over what water quality status arrives at the WWTP, and therefore, the efficiency metrics generally considered may not be sufficient to offer a fair comparison of WWTP performance. It is proposed that future exergy metrics should factor in the influent water status, for example, exergy destruction per kg COD removed.

EP

TE D

The exergy approach is promising; however, the requisite level of accurate data may pose challenges, primarily the availability of accurate energy, chemicals, flows, sludge outputs heavy metals, inter-process and emissions data, but also possible limitations in the current exergy approaches and models. Broader level monitoring, data availability and data integrity are necessary to assess the holistic environmental impact of wastewater treatment and this may need to be driven at the policy level. Should data quality improve, future work will reanalyse the WWTPs to assess the implications of the assumptions made and will be extended to cover sludge treatment and energy recovery.

AC C

With regard to the use of exergy analysis as a potential benchmarking approach and performance indicator, the exergy method is again promising. However, the inherent variability between WWTPs poses problems. For example, WWTPs may have different influent concentrations and flows, different discharge requirements including the requirement to remove nutrients, may use different technologies, and may be of widely differing scales inter alia. The exergy analyses presented here compared plants of similar scales, technologies and discharge requirements, but the results were very different as a result of different influent concentrations. One possible solution has been proposed and that is to consider various performance metrics for WWTP comparison, for example, the aforementioned exergy destruction per kg COD removed. One other way to overcome the difficulties associated with benchmarking may be to consider banding plants according to certain characteristics and benchmarking similarly banded plants. Future work will assess these options. A number of potential issues with the current approaches have been raised, primarily the reference or dead state environment choice, the general applicability of the chemical exergy

ACCEPTED MANUSCRIPT

RI PT

COD correlation to the various secondary treatment wastewater and sludge flows, and the need to account for water quality variations when comparing plant performance. The use of exergy analysis without corresponding consideration of water quality may lead to unfair comparisons between WWTPs and an additional metric was proposed to capture these variations. The REE metric highlights a number of important issues for policy makers, particularly in relation to increasing discharge requirements (i.e. lowering the discharge limits) and the additional energy required to treat wastewater to these more stringent limits. The thermodynamic efficiency of WWTPs is a function of both of these variables and will decrease explicitly as effluent standards become stricter and implicitly as the expected energy inputs increase.

TE D

M AN U

SC

6. CONCLUSIONS Exergy analyses of two Irish wastewater treatment plants were undertaken, using simultaneous energy audit and water quality measurement data, in an effort to compare plant performance and to assess the appropriateness of exergy analysis as a potential benchmarking approach and as a potential tool to influence policy. Although the plants were similar in scale, discharge requirements, and used similar technologies, the results of the wastewater treatment plant analyses differed significantly in terms of exergy destruction and exergetic efficiency. For example the rational exergetic efficient of WWTP E was found to be 27.5% in comparison to 40.2% for WWTP F. One important mitigating factor was the influent water quality differences between the WWTPs. Although not without current limitations, the approach proposed here is a new step in applying exergy analysis as a means to compare and potentially benchmark wastewater treatment plant performance that acknowledges the importance of simultaneous consideration of water quality to compare overall plant performance.

AC C

EP

ACKNOWLEDGEMENTS This report is published as part of the EPA Research Programme 2014-2020. The programme is financed by the Irish Government. It is administered on behalf of the Department of the Environment, Community and Local Government by the Environmental Protection Agency which has the statutory function of co-ordinating and promoting environmental research. The authors would like to acknowledge the EPA for the financial support. In addition, the authors would like to acknowledge the important input of Mr. Niall Durham to this research.

ACCEPTED MANUSCRIPT

BOD7 COD PQA RAS REE TIC TOC TOD TON TN TP WAS WWTP

Seven day Biochemical Oxygen Demand Chemical Oxygen Demand Power Quality Analyser Return Activated Sludge Rational Exergetic Efficiency Total Inorganic Carbon Total Organic Carbon Total Oxygen Demand Total Oxidised Nitrogen Total Nitrogen Total Phosphorous Waste Activated Sludge Wastewater Treatment Plant

Symbols α

Activity

Standard chemical exergy (kJ/mol unless otherwise stated)

TE D

e

Ch

SC

Conventional Activated Sludge Biochemical Oxygen Demand

M AN U

Acronyms AS BOD

RI PT

NOMENCLATURE

Standard Gibbs energy of formation (kJ/mol) Subscript indicates species under consideration Universal gas constant

T0 v x

Dead state temperature Stoichiometric coefficient Mole fraction

AC C

EP

∆Gf i R

7. REFERENCES

APHA, AWWA and WEF, 2005. Standard Methods for the Examination of Water and Wastewater. Cornelissen, R.L., 1997. Thermodynamics and Sustainable Development, PhD thesis, University of Twente. European Commission, 2013. Seventh Report on the Implementation of the Urban Waste Water Treatment Directive (91/271/EEC). Fitzsimons, L., 2011. A Detailed Study of Desalination Exergy Models and their Application to a Semiconductor Ultra-Pure water Plant, PhD thesis, Dublin City University.

ACCEPTED MANUSCRIPT Fitzsimons, L., Corcoran, B., Young, P., Foley, G., 2015. Exergy analysis of water purification and desalination: A study of exergy model approaches. Desalination, 212-224. Hellström, D., 2003. Exergy analysis of nutrient recovery processes. Water Science & Technology. 1, 27-36.

RI PT

Hellström, D., 1997. An exergy analysis for a wastewater treatment plant - An estimation of the consumption of physical resources. Water Environ. Res. 1, 44-51. Irish EPA, Wastewater Treatment Manuals: Primary, Secondary and Tertiary Treatment, 1997.

SC

Khosravi, S., Panjeshahi, M.H., Ataei, A., 2013. Application of exergy analysis for quantification and optimisation of the environmental performance in wastewater treatment plants. International Journal of Exergy. 1, 119-38.

M AN U

Kotas, T.J., 1995. The Exergy Method of Thermal Plant Analysis, 2nd ed. Krieger Publishing Company, Florida. Martínez, A., Uche, J., 2010. Chemical exergy assessment of organic matter in a water flow. Energy. 1, 77-84. Metcalf and Eddy, 2002. Wastewater Engineering: Treatment and Reuse, 4th ed. McGrawHill, New York, United States.

TE D

Olsson, G., 2012. Water and Energy: Threats and Opportunities. IWA Publishing, London. Roberts Alley, E., 2007. Water Quality Control Handbook, 2nd ed. McGraw-Hill, United States. Shannon, D., Byrne, N., Flynn, D., 2014. Focus on Urban Waste Water Treatment in 2013.

EP

Sharqawy, M.H., John, L.V., Zubair, S.M., 2011. On exergy calculations of seawater with applications in desalination systems. International Journal of Thermal Sciences. 2, 187-196.

AC C

Straub, C.P., 1989. Practical Handbook of Environmental Control. CRC Press, Florida, United States. Szargut, J., 2005. Exergy Method - Technological and Ecological Applications. WIT Press, Southampton, U.K. Szargut, J., Morris, D.R., Steward, F.R., 1988. Exergy Analysis of Thermal, Chemical and Metallurgical Processes. Hemisphere, New York. Szargut, J., Valero, A., Stanek, W., Valero, A., 2005. Towards an International Legal Reference Environment, Proceedings of ECOS, Trondheim, Norway. Szargut, J., 1989. Chemical exergies of the elements. Appl. Energy. 4, 269-286.

ACCEPTED MANUSCRIPT Tai, S., Matsushige, K., Goda, T., 1986. Chemical exergy of organic matter in wastewater. Int. J. Environ. Stud. 3-4, 301-315. Uche, J., 2000. Themoeconomic Analysis and Simulation of a Combined Power and Desalination Plant, PhD thesis, Universidad de Zaragoza. United States EPA, Wastewater Management Fact Sheet: Energy Conservation. 2006.

RI PT

Wagman, D.D., Evans, W.H., Parker, V.B., 1982. NBS Tables of Chemical Thermodynamic Properties: Selected values for inorganic and C1 and C2 organic substances in SI units. American Chemical Society and the American Institute of Physics for the National Bureau of Standards, Washington D.C.

AC C

EP

TE D

M AN U

SC

Yoshida, K., Nakayama, T., Yamauchi, S., 1984. Research on Socio‐Economic Aspects of Energy System, SPEY 4.

ACCEPTED MANUSCRIPT

APPENDIX A

WWTP E Day

02/09/2014

Tue

Dry

1833

1739

384.0

80.0

304.0

03/09/2014

Wed

Dry

1848

1606

480.0

138.7

341.3

63.8

04/09/2014

Thu

Dry

1795

1641

421.3

32.0

389.3

77.8

05/09/2014

Fri

Wet

1884

1742

455.0

144.0

311.0

06/09/2014

Sat

Wet

1879

1739

390.0

128.0

07/09/2014

Sun

Dry

1847

1709

1847.63

1696.00

426.07

104.53

Daily Average

Effluent (m3/day

COD in (mg/L)

WWTP E

COD out (mg/L)

COD removed (mg/L)

TN in (mg/L)

TP in (mg/L)

TP out (mg/L)

10.4

5.6

0.1

5.5

59.7

4.1

6.4

1.0

5.3

38.0

39.8

8.3

0.7

7.6

67.6

48.2

19.4

8.1

1.2

6.9

262.0

84.3

51.1

33.2

9.9

1.9

8.1

321.53

71.46

50.06

21.40

7.66

0.98

6.68

63.7

SC

Influent (m3/day)

M AN U

Weather

RI PT

Water Quality

Date

TN out (mg/L)

53.3

Energy

02/09/2014

Tue

03/09/2014

Wed

04/09/2014

Thu

05/09/2014

Fri

Weather

kWh/day

kWh/m

1866.01

1.02

Dry

1908.15

1.03

Dry

1897.93

1.06

Wet

1978.72

1.05

06/09/2014

Sat

Wet

1865.40

07/09/2014

Sun

Dry

1853.52

Daily Average

1894.96

kWh/kg COD removed

kWh/kg TSS removed

kWh/kg TN removed

kWh/kg TP removed

4.45

3.35

10.31

97.66

185.81

5.30

3.03

13.11

252.09

193.07

6.08

2.72

10.94

26.55

138.89

4.84

3.38

11.04

54.13

152.37

5.35

3.79

11.64

29.86

123.25

5.20

3.25

11.41

92.06

158.68

AC C

Dry

kWh/kg BOD removed

TE D

Day

EP

Date

3

0.99

1.00

1.03

TN removed (mg/L)

TP reduced (mg/L)

ACCEPTED MANUSCRIPT

WWTP F Influent (m3/day)

Effluent (m3/day

Tue

08-Oct

Wed

Dry

2705

2803

Wet

2021

2071

09-Oct

Thu

Dry

1830

1855

245.33

14-Oct

Tue

Dry

2053

2010

15-Oct

Wed

Wet

1802

16-Oct

Thu

Wet

19-Oct

Sun

Wet

Daily Average

COD in (mg/L)

COD out (mg/L)

COD removed (mg/L)

53.33

234.67

33.177

21.214

11.963

0.00

40.000

18.403

21.597

85.33

160.00

26.892

19.549

149.33

53.33

96.00

29.764

20.963

1742

256.00

48.00

208.00

25.067

0.000

1730

1586

341.33

117.33

224.00

23.984

9.690

14.294

1719

1543

192.00

32.00

160.00

28.320

22.163

1980.00

1944.29

245.33

64.89

154.67

29.60

16.00

kWh/day

kWh/m3

kWh/kg BOD removed

kWh/kg COD removed

kWh/kg TSS removed

kWh/kg TN removed

kWh/g TP removed

2.19

42.95

154.52

7.97

32.47

188.46

288.00

WWTP F

TN in (mg/L)

TN out (mg/L)

Day

07-Oct

Tue

Dry

1389.81

0.51

5.12

6.05

08-Oct

Wed

Wet

1417.24

0.70

8.40

09-Oct

Thu

Dry

1387.24

0.76

7.82

4.74

8.42

103.23

290.55

14-Oct

Tue

Dry

1466.22

0.71

6.28

7.44

7.78

81.15

256.62

15-Oct

Wed

Wet

1481.46

0.82

3.95

9.27

16-Oct

Thu

Wet

1562.65

4.03

10.01

63.19

430.74

19-Oct

Sun

Wet

1450.41

5.27

9.87

137.04

410.18

4.60

8.48

76.67

288.51

1450.72

AC C

Daily Average

EP

Date

TE D

Energy Weather

0.90

8.90

0.84

0.75

7.30

TN removed (mg/L)

RI PT

07-Oct

Water Quality Weather

SC

Day

M AN U

Date

TP in (mg/L)

TP out (mg/L)

TP reduced (mg/L)

5.037

1.712

3.325

4.454

0.733

3.721

7.343

2.994

0.385

2.609

8.801

3.589

0.806

2.783

2.915

0.818

2.097

6.157

2.717

0.660

2.057

11.69

3.63

0.85

2.77

3.683