High quality services of wastewater treatment require a continuous assessment and improvement of the technical, environmental and economic performance. This paper demonstrates a comprehensive approach for benchmarking wastewater treatment plants (WWTPs), using performance indicators (PIs) and indices (PXs), in a ‘plan-do-check-act’ cycle routine driven by objectives. The performance objectives herein illustrated were to diagnose the effectiveness and energy performance of an oxidation ditch WWTP. The PI and PX results demonstrated an effective and reliable oxidation ditch (good–excellent performance), and a non-reliable UV disinfection (unsatisfactory–excellent performance) related with influent transmittance and total suspended solids. The energy performance increased with the treated wastewater volume and was unsatisfactory below 50% of plant capacity utilization. The oxidation ditch aeration performed unsatisfactorily and represented 38% of the plant energy consumption. The results allowed diagnosing opportunities for improving the energy and economic performance considering the influent flows, temperature and concentrations, and for levering the WWTP performance to acceptable–good effectiveness, reliability and energy efficiency. Regarding the plant reliability for fecal coliforms, improvement of UV lamp maintenance and optimization of the UV dose applied and microscreen recommissioning were suggested.

INTRODUCTION

High quality services of wastewater treatment require a continuous assessment and improvement of the technical, environmental and economic performance. Nowadays, the scientific and technical community recognizes the benchmarking value as a tool for improving the performance through the systematic search and adaptation of leading practices (Cabrera et al. 2011). This definition, proposed by the specialist group of Benchmarking and Performance Assessment of the International Water Association, follows a ‘plan-do-check-act’ (PDCA) cycle routine, which involves two consecutive steps, performance assessment and performance improvement, and should always be driven by clearly defined objectives (Alegre et al. 2006).

In the past 20 years, many benchmarking projects have been undertaken in the water industry all over the world (Cabrera et al. 2011) and most European Union countries are assessing the performance of water and wastewater services regularly or irregularly, compulsorily or voluntarily. Since 2007, the European Benchmarking Co-operation (gathering with water utilities) offers an international benchmarking programme for water services aiming at facilitating water and wastewater utilities in the continuous process of improving and innovating water services and raising transparency (EBC 2014).

However, most performance assessment systems (PAS) apply to the undertaking as a whole and/or are particularly focused on the network systems (water supply and wastewater drainage) and economic funding issues (Stahre & Adamsson 2001; Matos et al. 2003; Ofwat 2004; Alegre et al. 2006; World Bank 2006; Stahre et al. 2008). Only a few (e.g. DWA 2008; Yang et al. 2010; Balmer & Hellström 2012) address the water and wastewater treatment systems, perhaps due to their wide variety which calls for treatment-specific performance metrics.

To fill this gap, LNEC – the Portuguese National Civil Engineering Laboratory – has been developing a PAS based on performance indicators (PIs) and indices (PXs). The Portuguese Initiative for the Performance Assessment of Water and Wastewater Treatment Plants (PASt21 project), involving 17 wastewater treatment plants (WWTPs), showed the PAS ability to assess the performance and to assist the continuous improvement of WWTP performance through benchmarking (Silva et al. 2012).

The objective of this paper is to demonstrate a comprehensive approach for diagnosing opportunities for improving the performance of a WWTP, using PIs and PXs in a PDCA cycle routine. For this purpose, the PDCA adapted for benchmarking WWTPs is described and the approach is illustrated for a municipal WWTP in the Algarve region (south of Portugal) for two key performance objectives: plant effectiveness and energy performance. The WWTP was designed for 44,500 inhabitants and includes an oxidation ditch and final UV disinfection.

PDCA cycle and methodology of PAS application for benchmarking WWTPs

The continuous improvement of WWTP performance should be implemented through PDCA cycles (Figure 1), which requires the verification and, eventually, the (re)definition of objectives and (re)selection of the corresponding assessment measures (PIs and PXs).
Figure 1

PDCA cycle routine adapted for benchmarking WWTPs.

Figure 1

PDCA cycle routine adapted for benchmarking WWTPs.

The PAS for WWTPs comprises: (i) a PI system for the overall assessment of the plant, on an annual-basis, in terms of treated water quality (Silva et al. 2014a), plant efficiency and reliability (Silva et al. 2014b), use of natural resources and raw materials with a special focus on energy (Silva & Rosa 2015), byproduct management (Silva 2016), safety, personnel, financial resources, and planning and design (Silva et al. 2012); and (ii) a system of PXs for assessing the daily performance in terms of treated water quality (Silva et al. 2014a), removal efficiencies (Silva et al. 2014b) and operating conditions (Silva 2016).

The PXs complement the PIs, since the PIs assess the overall performance of the WWTP in a given assessment period, whereas the PXs assess the distance to the goal and identifies ‘when’, ‘where’ and ‘why’ unsatisfactory, acceptable, good and excellent performances were obtained. The comprehensive integrated analysis of PIs and PXs (Figure 1) allows us to diagnose and to identify opportunities for improving the WWTP (technical, economic and environmental) performance.

The first step of PAS application is the definition of the objectives and of the assessment criteria for a given WWTP or a group of WWTPs. The PIs must then be selected and calculated accordingly, and analyzed against the pre-established reference values. Further insights may require the use of complementary PIs.

The next step is the selection and calculation of PXs of treated water quality to complement the information provided by the homologous PIs. The PIs assess the plant effectiveness in the assessment period (compliance with requirements – yes/no) and the PXs assess ‘when’ and how close or far each value was from the target.

One may then select the parameters and unit operations or processes for which one intends to assess and optimize the removal efficiency to enhance the treatment reliability. Removal efficiency is a mean to accomplish the treatment goals rather than a purpose and its relation with the plant economic and environmental efficiency is not straightforward. The novel methodology developed to translate removal efficiencies into PXs is detailed in Silva et al. (2014b).

The treated wastewater (TWW) quality and the removal efficiency criteria (PIs and PXs) will allow identifying the unit operations and processes that should be selected for assessing the operating conditions. In turn, the selected PXs of operating conditions will allow identifying ‘why’, ‘where’ and ‘when’ unsatisfactory, acceptable, good and excellent performances were obtained. Actions for improving the plant performance may then be proposed.

Case study

The WWTP is located in the east part of the Algarve region, in the south of Portugal. It was designed for 44,500 inhabitants, corresponding to 13,200 m3/day. The plant layout is illustrated in Figure 2, and includes screening, degritting/degreasing, two oxidation ditches with surface aerators, two secondary settlers, microscreening and UV disinfection to achieve 300 NMP/100 mL fecal coliforms in the final effluent. Unlike conventional oxidation ditches and extended aeration systems, these oxidation ditches were designed for 9–10 days of retention time. The sludge processing includes sludge rotary drum thickening and centrifuge dewatering (Figure 2). During the assessment period (2013–2015), only one oxidation ditch was working and the microscreens were decommissioned.
Figure 2

Layout of the WWTP.

Figure 2

Layout of the WWTP.

The study used a 3-year period (2013–2015) of regular monitoring data and a dedicated short-term campaign was carried out for energy use measurement, as detailed hereafter.

RESULTS AND DISCUSSION

The results of the PAS application are presented stepwise according to the PDCA cycle routine illustrated in Figure 1.

Step 1. Defining the objectives; selecting, computing and discussing the corresponding key PIs against the pre-established references values

As explained, the first step of PAS application was the definition of the objectives for this WWTP, which were the plant effectiveness and energy performance.

To address these objectives, seven key PIs were selected. The PI formulation and the reference values of each level of performance specific for this type of WWTP are presented in Table 1; the PI code and the existence of alternative options of a given PI (e.g. wtWQ03.1 and wtWQ03.2) are explained in Silva et al. (2014a). A green ball () represents ‘good’ performance, yellow () ‘acceptable’ and red () ‘unsatisfactory’ performances (please refer to the online version of this paper to see Table 1 in color: http://dx.doi.org/10.2166/wst.2016.432). The reference values were comprehensively derived in Silva et al. (2014a, 2014b) and Silva & Rosa (2015) and were based on a literature survey, empirical equations and/or field studies. Whenever adequate, the reference values incorporate the effect of key parameters responding to flow and concentration variations.

Table 1

Key PIs selected for each objective

Objective Key PIs Reference values
 
Good Acceptable Unsatisfactory 
Effectiveness (Silva et al. 2014a
 wtWQ01 – Quality tests carried out [%]
  Tests carried out (no.)/Tests required (no.) ×100 
 ≥100  [90; 100[  [0; 90[ 
 wtWQ02 – Parameters analyzed [%]
  Parameters analyzed (no.)/Parameters required (no.) × 100 
 ≥100    <100 
 wtWQ03.1 – Compliance of discharged wastewater quality with Directive 91/271/EEC [%]  ≥ 100      < 100 
 m = required parameters analyzed (no.)
Ji = compliance with parameter ‘i’ (0= no compliance or 1 = compliance) 
wtWQ03.2 – Compliance of discharged wastewater quality with DL 236/98 [%]  [95; 100]  [82; 95[  [0; 82[ 
 p = parameters analyzed (DL 236/98) (no.)
q = months assessed in the reference period
Jik = compliance with parameter ‘i’ in month ‘k’
(0 = no compliance or 1 = compliance) 
wtER01 – Volumetric efficiency [%]
 Treated wastewater (m3)/(Raw wastewater (m3) + Fresh water (m3))*100 
Energy performance (Silva & Rosa 2015)  
 wtRU03.1 – Energy consumption [kWh/m3]
  Energy consumption (kWh)/Treated wastewater (m3
 ≤0.280 + 1192/TW  ]0.280 + 1192/TW; 0.350 + 1,490/TW[  ≥0.350 + 1,490/TW 
 wtRU03.2 – Energy consumption [kWh/kg BOD5]
  Energy consumption (kWh)/BOD5 mass removed (kg) 
 ≤2  ]2; 10[  ≥10 
Objective Key PIs Reference values
 
Good Acceptable Unsatisfactory 
Effectiveness (Silva et al. 2014a
 wtWQ01 – Quality tests carried out [%]
  Tests carried out (no.)/Tests required (no.) ×100 
 ≥100  [90; 100[  [0; 90[ 
 wtWQ02 – Parameters analyzed [%]
  Parameters analyzed (no.)/Parameters required (no.) × 100 
 ≥100    <100 
 wtWQ03.1 – Compliance of discharged wastewater quality with Directive 91/271/EEC [%]  ≥ 100      < 100 
 m = required parameters analyzed (no.)
Ji = compliance with parameter ‘i’ (0= no compliance or 1 = compliance) 
wtWQ03.2 – Compliance of discharged wastewater quality with DL 236/98 [%]  [95; 100]  [82; 95[  [0; 82[ 
 p = parameters analyzed (DL 236/98) (no.)
q = months assessed in the reference period
Jik = compliance with parameter ‘i’ in month ‘k’
(0 = no compliance or 1 = compliance) 
wtER01 – Volumetric efficiency [%]
 Treated wastewater (m3)/(Raw wastewater (m3) + Fresh water (m3))*100 
Energy performance (Silva & Rosa 2015)  
 wtRU03.1 – Energy consumption [kWh/m3]
  Energy consumption (kWh)/Treated wastewater (m3
 ≤0.280 + 1192/TW  ]0.280 + 1192/TW; 0.350 + 1,490/TW[  ≥0.350 + 1,490/TW 
 wtRU03.2 – Energy consumption [kWh/kg BOD5]
  Energy consumption (kWh)/BOD5 mass removed (kg) 
 ≤2  ]2; 10[  ≥10 

For plant effectiveness, the PIs assess the plant performance regarding the wastewater quality tests carried out (wtWQ01), the parameters analyzed (wtWQ02), the discharged wastewater quality compliance with Directive 91/271/EEC for parameters BOD5 (biochemical oxygen demand), COD (chemical oxygen demand) and TSS (total suspended solids) (wtWQ03.1) and with the Portuguese decree-law (DL) 236/98 for fecal coliforms (wtWQ03.2) and the volumetric efficiency (wtER01).

For the energy performance, one key PI was selected with two options (explained in Silva & Rosa (2015)), i.e. the energy consumption per cubic meter (wtRU03.1) and per BOD5 mass removed (wtRU03.2). The WWTP contains no anaerobic digestion, so the PIs relative to the net use of energy from external sources (wtER08) and to the production of energy from biogas (wtBP18) (Silva & Rosa 2015) were not applicable.

The selected PIs were computed for a 3-year period (2013–2015) and compared with the reference values derived in the earlier studies and presented in Table 1. The PI values are presented in the supplementary section (available with the online version of this paper) and the results of the PI judgment for each objective are presented in ‘donut’ graphics (Figure 3). Each part of the donut represents a key PI, where a green slice stands for ‘good’ performance, a yellow for ‘acceptable’ and a red for ‘unsatisfactory’ performance (please refer to the online version of this paper to see Figure 3 in color: http://dx.doi.org/10.2166/wst.2016.432).
Figure 3

Overview of the WWTP performance in each objective (PIs in Table 1).

Figure 3

Overview of the WWTP performance in each objective (PIs in Table 1).

Relative to plant effectiveness, the performance was constant in the 3 years and good relative to volumetric efficiency, wastewater quality tests carried out, parameters analyzed and the compliance with Directive 91/271/EEC. Concerning the quality compliance with the fecal coliform requirements (wtWQ03.2), the performance was unsatisfactory during the analyzed period (75% in 2013–2014 and 60% in 2015).

The energy performance in terms of energy consumption per cubic meter varied from unsatisfactory performance in 2013 to acceptable performance in 2014 and again unsatisfactory performance in 2015 and was always acceptable in terms of energy consumption per BOD5 mass removed. One should note that, taken into consideration the data available (more limited for BOD5 mass removed, particularly for smaller plants, with less intensive monitoring), the reference values derived for the energy consumption per BOD5 mass removed are less exigent than those of the energy consumption per volume treated, particularly in the high range (Silva & Rosa 2015).

Step 2. Selecting and computing the relevant complementary PIs; discussing the results of key and complementary PIs

In addition to the key PI results, complementary PIs may provide helpful insights and the same applies to the analysis of the daily PI results and their relation with other variables.

Related to effectiveness, no complementary PIs were selected, and this objective will be further analyzed by PXs.

For energy performance, the complementary PIs selected were the adequacy of plant capacity (wtER13) and BOD5 capacity (wtER15). An adequate capacity means that the plant is more than 60% (or >80% for good performance) of the time (during the assessed year) using 70–95% of its capacity, with the lower limit corrected to accommodate the seasonality (Silva et al. 2013). The PI results show that the WWTP was oversized during the assessment period. Analyzing the daily flows and comparing with the 70–95% range of plant capacity, one can see the WWTP underutilization (Figure S1 in supplementary section, available with the online version of this paper), even in the high season, i.e. the period of 30 consecutive days with the highest average (30 max). This WWTP shows a wide flow variation without the summer seasonality commonly found in the Algarve. Actually, the drainage area of this WWTP has no touristic peak activity during the summer. Instead, a flow variation pattern was found with the day of the week, i.e. in the 3 years analyzed, the box-plots were similar from Monday to Wednesday, from Thursday to Friday, and during the weekend; the median and average values were lower in Thursday to Friday and higher during the weekend (Figure S2 in supplementary section, available with the online version of this paper).

To assess the impact of the flow variation on the energy performance, the wtER03.1 PI was computed on a daily basis and was related with the treated wastewater volume (Figure 4).
Figure 4

PI wtER03.1 vs. volume treated and wtER03.2 vs. BOD5 mass removed ( good acceptable and unsatisfactory performances).

Figure 4

PI wtER03.1 vs. volume treated and wtER03.2 vs. BOD5 mass removed ( good acceptable and unsatisfactory performances).

As expected, Figure 4 (left) shows that the unit energy consumption was lower for the days with higher treated volumes. Actually, when comparing WWTPs of the same technology and adequate capacity utilization (nor over- nor underutilized) (WERF 2011) there is an inverse relation between the unit energy consumption and the plant's size-scale (expressed by the treated wastewater volume), as accounted by the reference values in Table 1 – bigger plants use the energy more efficiently. For the same WWTP, different wastewater volumes treated may also affect the plant energy performance, since the plant capacity utilization changes and the closer the WWTP is to its design capacity, the more efficient is the unit energy consumption (WERF 2011; Silva & Rosa 2015).

Figure 4 shows that below 3,200 m3/d, the plant energy performance was always unsatisfactory (red dots), i.e. when the WWTP was underutilized below 50% of its capacity, the plant was not able to reduce the energy consumption according to the flow (volume or mass) reduction, so the energy used exceeded the energy required for treating the wastewater to the desired limit values. For 3,200–5,500 m3/d the plant energy performance was unsatisfactory or acceptable (red and yellow dots). (Please refer to the online version of this paper to see Figure 4 in color: http://dx.doi.org/10.2166/wst.2016.432.) In turn, when higher wastewater volumes were treated (or higher mass was removed), the unit energy consumption was lower, corresponding to acceptable-good performance (5,500–7,000 m3/d) or mostly to good performance for treated wastewater volumes above 7,000 m3/d (Figure 4, left).

The same behavior was found for the daily energy consumption per mass removed (wtRU03.2) vs. the BOD5 mass removed (2,400 kg/d design capacity; Figure 4 right) – acceptable performance below 60% plant capacity and good performance above 67% (1,600 kg/d).

The 3-year data available in this study showed a decreasing linear relation between kWh/m3 and kWh/kg BOD5 and varied in the range 2.5:1 to 6:1. The 3.4:1 ratio was found in field studies and values above this trend line (i.e. higher energy consumption per mass removed) may indicate dilute (e.g. stormwater) inflows whereas values below may alert for industrial (highly charged) inflows.

Step 3. Selecting, computing and discussing the relevant PXs of treated wastewater quality

The earlier PI results alert to the aspects to be further analyzed in the operational performance assessment through the PXs, aiming at identifying ‘why’, ‘where’ and ‘when’ unsatisfactory, acceptable, good and excellent performances were obtained. The reference values used in the PXs are presented in the supplementary section (Table S2, available with the online version of this paper).

The next step in PAS application is the selection and calculation of the PXs of treated water quality to complement the information provided by the homologous PIs. The PIs assessed the plant effectiveness (yes/no) and showed a good performance for BOD5, COD and TSS parameters and unsatisfactory for fecal coliforms. To assess and improve the daily operational performance of the WWTP in terms of treated water quality, PXs of BOD5, COD and TSS and fecal coliforms were computed (Figure 5).
Figure 5

PXs of BOD5, COD, TSS and fecal coliforms in the treated wastewater.

Figure 5

PXs of BOD5, COD, TSS and fecal coliforms in the treated wastewater.

The PXs show the distance that remains to achieve the pre-established objectives, from which safety margins and improvement opportunities (technical but also economic) may be derived. In this case, economic gains may be foreseen for BOD5, COD and TSS, since the performance was always good or excellent (Figure 5). Related to the fecal coliforms, the PI identified non-compliance and the PX showed a huge variation from excellent to unsatisfactory performance during the assessment period. To better understand these variations, the PXs of operating conditions were selected in Step 5.

To assess the plant reliability, the statistical analysis proposed in Silva et al. (2014a) was applied. In addition to the PX percentage in each performance level (pie chart), Figure 6 shows the percentile distribution of treated wastewater PXs. The flatter the profile is (this is, the closest the percentiles 25 (P25) and 75 (P75) are from the median) the better is the plant reliability. If the median corresponds to acceptable performance, a flat profile indicates a reliable effectiveness of the plant.
Figure 6

Long-term operational performance levels of TWW quality in the 3-year period.

Figure 6

Long-term operational performance levels of TWW quality in the 3-year period.

The profiles of BOD5, COD and TSS (always in the green region – good performance, Figure 6) indicate opportunities for improving the plant economic performance if lowering the performance level results in cost savings, in any case keeping the performance in the acceptable–good (yellow–green) region. (Please refer to the online version of this paper to see Figure 6 in color: http://dx.doi.org/10.2166/wst.2016.432.)

In the case of fecal coliforms, the percentile distribution profile shows that more than 60% of the values presented excellent performance, but the plant reliability was low since the slope of the profile was high (Figure 6).

Step 4. Selecting, computing and discussing the relevant PXs of removal efficiencies

In this application, the PXs of removal efficiencies were not selected, since in this WWTP there is a single barrier for organics (BOD5 and COD) and suspended matter (TSS). As so, the wastewater quality indices can directly demonstrate the good performance of the oxidation ditch and a non-reliable UV disinfection.

Step 5. Selecting, computing and discussing the relevant PXs of operating conditions

At this phase, the operating conditions indices related to the objectives defined should be selected. Related to effectiveness, the operating conditions of UV disinfection were selected to analyze the unstable performance towards the fecal coliforms. According to the available data, the indices of influent transmittance and TSS were selected. The transmittance results showed a decreasing performance from early 2014 to mid-2015, with some episodes of unsatisfactory performance. This trend may have influenced the fecal coliforms' performance in the treated wastewater (TWW), as shown in Figure 7.
Figure 7

PXs of transmittance in UV disinfection and comparison with PXs of fecal coliforms in treated wastewater and the TSS influent to the UV.

Figure 7

PXs of transmittance in UV disinfection and comparison with PXs of fecal coliforms in treated wastewater and the TSS influent to the UV.

The lower performance of transmittance presents a slight relation with the TSS peaks (low peaks, 4–14 mg/L) to the UV disinfection (Figure 7). Measures to address this issue may include the verification and eventual improvement/adjustment of the lamp maintenance practices and optimization of the UV dose applied and the recommissioning of the microscreens prior to the UV disinfection (in place but not in operation). The most cost-effective measure(s) should be identified by the utility and implemented.

Related to the energy performance, the PXs of each use in biological treatment were selected, as well as the operating conditions that influence the energy consumption. The reference values for the energy PXs relative to aeration, recirculation and sludge wasting in the 3-year period were computed.

Concerning the aeration, the reference values of energy consumption (EV O2 in Silva 2016) were calculated based on the BOD5 influent concentration to the reactor and on the oxygen transferred under field conditions and the biomass wasted. The biomass wasted varies with the mixed-liquor suspended solids (MLSS), the detention time and the solids retention time (SRT). The PXs of MLSS (Figure 8) show a good performance and an unsatisfactory performance during the first semester of 2014, corresponding to excessive MLSS, threefold the maximum acceptable in oxidation ditches (4,000 mg/L).
Figure 8

PXs of MLSS in the oxidation ditch.

Figure 8

PXs of MLSS in the oxidation ditch.

The detention time exhibits in some days (15%) an unsatisfactory performance due to insufficient detention times (below 18 h). The SRT also showed unsatisfactory performance associated with values below the minimum acceptable for oxidation ditches (20 days). However, this WWTP was designed for retention times between 9 and 11 days and the observed retention times of 10 days on average were effective, corresponding to good performance of wastewater quality. This operation mode, with lower SRTs, has lower energy demand for aeration, as shown by the reference values in Figure 9 left vs. right.
Figure 9

Reference values of energy consumption for aeration with typical SRT in the oxidation ditch and with real SRT (computed with daily data).

Figure 9

Reference values of energy consumption for aeration with typical SRT in the oxidation ditch and with real SRT (computed with daily data).

The energy required for sludge wasting also depends on the SRT and, in this case, the lower retention times require more energy for sludge wasting, though with a low impact on the WWTP energy consumption (reference values of 0.2–9 kWh/m3 vs. 0.2–5 kWh/m3, as illustrated in Figure S3, available with the online version of this paper).

The reference values of energy consumption in recirculation were based on the typical return sludge ratio in oxidation ditch for C and N removal, where performance varied from unsatisfactory to good, the 39% unsatisfactory values being due to excessive sludge return ratios (Figure 10).
Figure 10

PXs of sludge return ratio and the long-term operational performance levels in the 3-year period.

Figure 10

PXs of sludge return ratio and the long-term operational performance levels in the 3-year period.

For improving the WWTP energy performance, one needs to measure and collect the energy actually consumed in each use in the WWTP and to compare it with the reference values computed. For this purpose, an energy campaign was carried out to measure the energy consumption of each use in the biological treatment. The equipment analyzed were: the three aerators, one with variable speed and one with a progressive starter, both working with dissolved oxygen set points, and one working by timing (5 min on and 120 min off); three mixers; two recirculation pumps and two sludge wasting pumps. A power energy logger PEL 103 was used to measure the equipment power and energy consumption. The aeration represented 38.1% of total energy consumption in the WWTP, where the surface aerator 1 corresponded to 22.9% (Figure 11). In this campaign period (1–3 December 2015) the energy consumption was 0.79 kWh/m3 for 4,256 m3/d treated wastewater, which corresponded to unsatisfactory performance, as observed in 2015 for the energy consumption PI (Figure 3 for annual energy consumption and Figure 4 for daily energy consumption).
Figure 11

Weight of each energy use.

Figure 11

Weight of each energy use.

Based on the campaign results, the PXs of energy consumption in each use could be computed, as well as their reference values for the campaign week. The values measured, the reference values and the PX results of energy consumption are presented in Table 2. The PX results showed unsatisfactory performance (high energy consumption compared to the reference values) in aeration and sludge wasting during the campaign. Aeration represented 38.1% of the plant overall energy consumption (Figure 11), identifying where the potential for improvement exists.

Table 2

PXs of energy consumption

    Reference values
 
  
Energy consumption in… Value measured R200 R100 R PX 
Aeration 300 Wh/m3 162 244 325 30  
Mixing 9.59 Wh/m3 6.7 10.9 13.7 132  
Recirculation 11.9 Wh/m3 7.7 23 36.9 172  
Sludge wasting 3.19 Wh/m3 0.6 1.9 3.5 20  
    Reference values
 
  
Energy consumption in… Value measured R200 R100 R PX 
Aeration 300 Wh/m3 162 244 325 30  
Mixing 9.59 Wh/m3 6.7 10.9 13.7 132  
Recirculation 11.9 Wh/m3 7.7 23 36.9 172  
Sludge wasting 3.19 Wh/m3 0.6 1.9 3.5 20  

Step 6. Identifying improving opportunities

The next step would be the identification of the opportunities for improving the performance.

Regarding the plant reliability for fecal coliforms, the improvement of UV lamp maintenance and optimization of UV dose applied and the microscreens' recommissioning are suggested.

Another improvement measure identified was to adjust the aeration to the influent flows, temperature and concentrations; another was levering the performance to acceptable–good effectiveness and energy performance (decreasing the former from excellent to good, and improving the latter).

Step 7. (Re)defining the objectives and start another cycle (1–6)

Having arrived at this point, the utility should proceed to the next PDCA cycle, (re)defining the objectives and (re)selecting the corresponding assessment measures (PIs and PXs) according to the diagnosis made in the last cycle, as suggested in the conclusions.

CONCLUSIONS

This paper demonstrated a comprehensive approach for diagnosing opportunities for improving the performance of a WWTP using PIs and PXs in a PDCA cycle routine. This approach was illustrated stepwise for an oxidation ditch WWTP with final UV disinfection.

Since this was the first PAS application to this WWTP, the performance objectives were to diagnose the plant effectiveness and energy performance. The selected PIs and PXs were computed for a 3-year period (2013–2015) and the reference values proposed in earlier studies for these water and sludge processing units were used (which, whenever relevant, consider the influent characteristics).

The PIs and PXs of the plant effectiveness showed a good–excellent performance for BOD5, COD and TSS parameters, and an overall unsatisfactory performance for fecal coliforms (no compliance shown by the PI) though varying from unsatisfactory to excellent depending on the daily influent transmittance and TSS (PX results). These results demonstrated an effective and reliable oxidation ditch and a non-reliable UV disinfection. Regarding this, the improvement of UV lamp maintenance and optimization of UV dose applied and the microscreens' recommissioning were suggested.

The energy performance increased with the treated wastewater volume, showing a relation with the plant capacity utilization. The closer the WWTP was to its design capacity, the more efficient was the unit energy consumption, and below 50% of plant capacity utilization, the energy performance was unsatisfactory. In this case, as the WWTP was already working with one oxidation ditch, no improvements were identified. Going deeper into the processes through the PXs, the energy consumption in aeration performed unsatisfactorily and represented 38% of the total energy consumption in the plant. These results pointed out the advantages of undertaking seasonal measures for improving the energy and economic performance by adjusting the aeration to the influent flows, temperature and concentrations. Actually, the PXs of wastewater quality identified room for economic improvement, since the BOD5, COD and TSS performance was always good or excellent. The water utility should therefore verify if lowering the oxidation ditch effectiveness performance to an acceptable–good level does improve the energy performance to an also acceptable–good level with subsequent cost savings.

At this point, based on the improvement opportunities diagnosed, the water utility should define the action plan, including the measures to be implemented for achieving the new objectives, which should consider quantitative goals for improvement. For instance, in this case, this could be to leverage the BOD5, COD and TSS performance as well as the energy performance to indices of 150 (100 minimum acceptable, 200 good performance). New objectives may require new metrics and this selection should be as rational as possible. The full portfolio of PIs and PXs is very broad and is not to be calculated in every application, plant and assessment period. Instead, as explained, from the PI and PX system one should select a set of PIs and PXs according to the assessment objectives for a given WWTP and year.

ACKNOWLEDGEMENTS

This research has received funding from the Portuguese Science and Technology Foundation (C. Silva's PhD scholarship, SFRH/BD/80295/2011) and from European Union LIFE Programme under the agreement LIFE14 ENV/PT/000739 (LIFE IMPETUS). The publication reflects only the authors' views and European Union is not liable for any use that may be made of the information contained therein. The authors further acknowledge the colleagues of Águas do Algarve for providing the data of the WWTP analyzed and for carrying out the energy campaign.

REFERENCES

REFERENCES
Alegre
H.
Melo Baptista
J.
Cabrera
E.
Jr
Cubillo
F.
Duarte
P.
Hirner
W.
Merkel
W.
Parena
R.
2006
Performance Indicators for Water Supply Services
.
Manual of Best Practices Series
,
2nd edn
.
IWA Publishing
,
London, UK
.
Balmer
P.
Hellström
D.
2012
Performance indicators for wastewater treatment plants
.
Water Science and Technology
65
(
7
),
1304
1310
.
Cabrera
E.
Jr
Dane
P.
Haskins
S.
Theuretzbacher-Fritz
H.
2011
Benchmarking Water Services. Guiding Water Utilities to Excellence
.
IWA Publishing
,
London, UK
.
Co-published by AWWA
.
DWA
2008
Corporate Benchmarking – Metric Benchmarking as Component of the Modernisation Strategy – Performance Indicators and Evaluation Principles
.
DWA German Association for Water, Wastewater and Waste (A. Schulz as Spokesperson of the working group)
,
Germany
.
EBC
2014
PROJECT PLAN, IB2013 – Benchmarking Exercise
.
European Benchmarking Co-operation
,
The Netherlands
,
Danish Water and Wastewater Association (DANVA), Norwegian Water (Norsk Vann), Finnish Water Utility Association (FIWA) and Association of Dutch Water Companies (Vewin). Supported by European Federation of National Water Utility Associations (EurEau) and International Water Association (IWA)
.
Matos
R.
Cardoso
A.
Ashley
R.
Duarte
P.
Molinari
A.
Schulz
A.
2003
Performance Indicators for Wastewater Services
.
Manual of Best Practices Series
.
IWA Publishing
,
London, UK
.
Ofwat
2004
Updating the Overall Performance Assessment (OPA) – Conclusions and Methodology for 2004–05 Onwards
.
UK Office of Water Services
,
Birmingham
,
UK
.
Silva
C.
2016
Assessing and Improving Wastewater Treatment Performance. A Contribution to the 3rd Generation of the Assessment System
.
Doutoramento em Engenharia do Ambiente. Universidade de Lisboa, Instituto Superior Técnico
,
Lisboa, Portugal
.
Silva
C.
Ramalho
P.
Quadros
S.
Alegre
H.
Rosa
M. J.
2012
Results of ‘PASt21’ – the Portuguese initiative for performance assessment of water and wastewater treatment plants
.
Water Science and Technology: Water Supply
12
(
3
),
372
386
.
Silva
C.
Ramalho
P.
Rosa
M. J.
2013
PAS application in Portugal. In Pilot-city specific, new technological options for stormwater separation and optimized WWTP, Part I. Curative actions in wastewater treatment systems
.
D43.1 Report. TRUST EU project – Transitions to the Urban Water Services of Tomorrow
.
Silva
C.
Quadros
S.
Ramalho
P.
Rosa
M. J.
2014a
A tool for assessing treated wastewater quality in urban WWTPs
.
Journal of Environmental Management
146
,
400
406
.
Silva
C.
Quadros
S.
Ramalho
P.
Alegre
H.
Rosa
M. J.
2014b
Translating removal efficiencies into operational performance indices of wastewater treatment plants
.
Water Research
57
,
202
214
.
Stahre
P.
Adamsson
J.
2001
Performance benchmarking: a powerful management instrument for water and wastewater utilities
.
Water Technology
12
,
47
77
.
Stahre
P.
Adamsson
J.
Mellstrom
G.
2008
A new approach for assessment of the performance of water distribution and sewerage networks
. In:
International Conference on Performance Assessment of Urban Infrastructure Services, Drinking Water, Wastewater and Solid Waste
(
Cabrera
E.
Pardo
M. A.
Jr
, eds).
IWA Publishing, London, UK
.
WERF
2011
Energy Management. Exploratory Team Report Executive Summary
.
Water Environment Research Foundation
,
Alexandria, VA, USA
.
World Bank
2006
IBNET Indicator Definitions
.
World Bank
,
Washington, DC, USA
.
Yang
L.
Zeng
S.
Chen
J.
He
M.
Yang
W.
2010
Operational energy performance assessment system of municipal wastewater treatment plants
.
Water Science and Technology
62
(
6
),
1361
1370
.

Supplementary data