3.5.1. Technical Process Assessment (Gate II)
As described above, the shift from WWTPs towards WRFs requires that resource recovery technology integration becomes a central objective in process design; hence, technical decisions need to be guided to meet both treatment and recovery requirements. Therefore, the technical assessment will assess the performance of a process in the dimensions of (i) treatment, (ii) operation (
Table 4), and (iii) resource recovery (
Table 5). The MEBs conducted in step 3 provide the basis for estimating various technical performance criteria that are presented in
Table 4 and
Table 5. The treatment performance of new processes is usually assessed by the removal efficiencies of pollutants that define the legal effluent quality, like, e.g., COD, TKN, and total phosphorous (TP) [
35]. Since the environmentally safe release of treated water into surface water bodies remains the primary goal of designed processes, each process design has to be assessed accordingly. It is possible that effluent quality requirements will become stricter and ask for more indicator substances in the future. If that is expected, the treatment performance assessment can be extended by estimating the capability of a process to fulfil potential future legal requirements for micro-pollutants [
43] and/or nutrient removal [
32].
The integration of resource recovery technologies into treatment plants can imply operational uncertainties [
26], and therefore the question about which technologies are most useful and how to combine them in process design has to be tackled [
90]. Therefore, in addition to the treatment performance, the operation of each process design can be assessed. There have been criteria proposed and established for treatment process operation assessment which can be applied here [
27,
49,
53]. Operational data and information for single process units, like, e.g., sedimentation tanks or bioreactors, can be extracted from the vast literature available on particular treatment technology operations [
9,
29,
30]. Operational data for more innovative resource recovery units can be extracted from specific articles describing pilot or case studies. Searching the term “wastewater resource recovery technology” in an online search engine for scientific publications showed over 243,000 results in 2020.
After assessing the treatment performance and operation, each process can be assessed regarding its resource recovery performance, which includes criteria to assess the expectable quality of recovered resources on the one hand and the recovery efficiency of the process on the other hand. The end-of-waste concept proposes to facilitate recycling by defining that waste is no longer perceived as such if it has undergone a recovery process that ensures the use of a recovered product will not lead to overall adverse human health impacts [
25]. Several studies have shown that resources recovered from municipal wastewater may be of uncertain quality or even contaminated, which may impose health risks. For example, struvite has been reported to possibly contain heavy metals [
91], and reclaimed water may contain harmful by-products of chemical biocides used in tertiary water treatment [
92]. Studies examining health risks in the field of circular economy seem mostly to deal with occupational health risks which relate to the workplace and less with health risks arising from recovered product use. In the context of WRFs, the only resource that has been subject to extensive risk management considerations is reclaimed water. Risks and therefore quality requirements for reclaimed effluents depend on the intended reuse type. Biological and human health safety control measures need to be proactively developed as legal standards are often missing for each reuse type [
75]. Major concerns are pathogens that can cause acute infections at very low doses upon exposure, but also chemical micro-pollutants need to be removed from the water for safe reuse [
93]. A systemic risk management approach that covers all aspects of the reclaimed water production, distribution, and utilization has been proposed but needs further elaboration and a proactive management approach [
73,
94]. To resolve legal uncertainties, the European Commission recently proposed minimum risk control standards for water reuse that can provide guidance for designing advanced treatment processes for safe-to-use reclaimed water [
95]. The quality of a recovered resource should not only be safe for human health but also competitive with conventional alternatives on the market. After the potential applications and customers associated with a process have been analyzed in step 4, it is now important to assess whether a process can cope with the quality requirements of both. This can be a challenge, as it is one thing to recover a product from wastewater but another to obtain a marketable quality from the recovery process. For example, it is technically feasible to recover biochemicals like, e.g., volatile fatty acids (VFAs) or PHA from COD, but to obtain a certain purity requested from the industry is a technical challenge in mixed culture systems [
8] and thus has to be considered from the early design stage on.
Next to resource quality, the resource recovery performance of a process can be assessed in terms of recovery efficiency. The mass efficiency of process designs can be assessed with the results of the mass and energy balances conducted in step 3. For example, if two different processes recover P but process (A) integrates struvite crystallization [
96] while process (B) integrates the magnetic extraction of vivianite [
97], it is useful to assess the recovery rate of the influent P of each process. Furthermore, if resource recovery technologies include not only the extraction but also on-site refinement of a recovered product, KPIs originally developed for the chemical or pharmaceutical process assessment could be applied to assess the performance of a particular recovery technology in comparison to an alternative one. For example, the chemical extraction of extracellular polymeric substances (EPS) from aerobic granular sludge is possible with different solvents [
98] and could be assessed by the solvent score method [
71].
3.5.2. Cost–Benefit Analysis (Gate III)
After the technical assessment has been completed and processes with a technically unsatisfying performance have been either re-designed for improved performance or discarded as an unfeasible alternative, the economic performance of technically promising designs needs to be assessed. The European water framework directive demands that urban water systems should be economically self-sustained, meaning that costs should be covered by the system itself through water pricing and service fees for wastewater treatment [
15]. Taking the latter into account, a WRF has a different “business model” compared to conventional WWTPs. Although its priority remains to treat wastewater and charge a fee for this environmental and human health service provision, a WRF ideally also generates revenues from recovered resource sales [
90]. Resource recovery introduces new financial uncertainties and leads to a whole different cost–benefit structure within the wastewater treatment equation [
22,
37,
103]. Not only are additional investment costs likely to occur from required recovery units and installations, but also substantial changes in operating costs can be expected if processes are designed from a resource recovery perspective. For example, integrating chemically enhanced primary treatment into a process that uses anaerobic sludge stabilization may alter the process economics in various ways. On the one hand, it promises economic benefits from increased methane yields and decreased aeration requirements for the biological treatment. On the other hand, it requires the usage of special polymers that represent an additional cost factor [
67]. Another example would be the reclamation of reusable water accomplished by advanced membrane-based treatment. The additional treatment step leads on the one hand to additional operational costs for energy consumption [
44], maintenance costs to prevent membrane fouling [
104], and waste management costs for retentate handling [
105]; on the other hand, it may generate steady economic benefits in the form of revenues from water customers [
106]. Furthermore, P recovery requires additional investment costs but may also lead to lower waste management costs due to improved sludge dewaterability and benefits from recovered product marketing [
97].
These examples show clearly that a reliable statement about the economic feasibility of a WRF can only be made if its combined costs and benefits are assessed. Therefore, this step aims to conduct a cost and benefit analysis which is a systematic approach for estimating and comparing the positive and negative economic consequences of an investment to determine its net profitability. It follows the simple formula (1):
where
NP is net profit,
B is the benefits of item
i, and
C is the costs of item
i [
34].
To assess the economic performance of processes, the time horizon that the new process is expected to be operated in needs to be defined. A 20-year time horizon to calculate the costs and benefits associated with wastewater resource recovery processes has been suggested [
61,
103], but it may be defined more accurately in accordance with site and project-specific circumstances. The economic performance of different process designs can only be compared if the same time horizon is applied in the assessment so that they have the same time to accumulate costs and benefits [
107]. The costs and benefits that will arise during the expected time horizon are calculated for each process using the net present value (NPV). The NPV expresses the monetary value of future cash flows and is an indicator to determine the economic value of a process design and thus allows the ranking of alternatives. The higher the NPV, the more economically favorable a process design is. The NPV method requires the determination of a discount rate that accounts for the opportunity costs of time by discounting future costs and benefits because of the profit that could be earned in alternative investments. It is a widespread practice in CBA to use the current market interest rate as the discount rate [
107]. The discount rate calculation is shown in formula (2). To calculate the NPV, the time horizon is usually divided into yearly periods and the net profits are discounted and calculated on a yearly basis, which leads to formula (3).
where the discount rate r equals the present value of 1€ received in n years when the interest rate i is compounded annually [
107].
where the net present value (NPV) at time t, calculated for a time horizon of n years, is the sum of discounted annual net profits (NP), assuming a discount rate i. Adapted from [
107].
Table 6 shows the cost and benefit factors to be included in CBA for a WRF process. The method required that all cost and benefit factors related to a process design are estimated for each year of the time horizon. The annual net benefits then need to be discounted and summed up to obtain the NPV of the process. To avoid the need of predicting future price level changes by, e.g., inflation, the costs and benefits should be expressed in current real prices and not with nominal prices [
108]. Many occurring cost and benefit factors can be deduced from market prices or estimated using literature studies or expert judgement [
64].
As in most infrastructure projects, a high initial total investment cost occurs at the beginning of the time horizon when the new process is planned, purchased, and constructed. A residual value of the fixed investments must be included within the investment costs occurring in the last year if the plant is believed not to be liquidated after the time horizon has ended. It reflects the capacity of the remaining service potential of fixed assets which are not yet completely exhausted [
108]. The operational costs of wastewater treatment are usually measured on the basis of contaminant removal [
33], which requires taking the influent characteristics into account that have been defined in step 1. Some cost factors may require general assumptions to keep a certain degree of simplicity during this early stage economic assessment. For example, the electricity needed to supply oxygen into biological treatment units depends on the saturation concentration of oxygen at an assumed temperature, pressure, and salinity [
11]. Expectable benefits from resource sales can be calculated based on the quantities of recoverable resources that have been estimated for each process design in step 3 by MEBs and their market prices analyzed in step 4. In addition to resource sales, benefits could also be gained from charges for handling additional waste streams as explained in step 4, like, e.g., food wastes.
Costs and benefits are often defined in CBA as decreases and increases in human wellbeing, which can include various external effects of human behavior that have no market prize [
34,
51]. Although the monetization of these effects can be achieved by different methods [
107], the cost and benefit factors suggested in
Table 6 do not account for external effects like, e.g., the cost of undesirable effluent constituents entering surface water bodies. The reason is that the monetization of external effects is a complicated procedure usually conducted by experts like, e.g., environmental economists. This framework is supposed to be useful for institutions and decision makers in the wastewater sector that do not necessarily have a strong expertise in those specializations. The environmental impacts of a process will be carefully assessed in the next assessment stage. After the CBA has been completed for each process design, the results can be reconciled with the budget estimation conducted in step 1 to identify any variances between the available budget and the estimated costs of a process.
3.5.3. Environmental Impact Assessment (Gate IV)
Finally, after the technical and economic assessment has been completed and those process designs that did not perform well in one or both dimensions have been either re-designed for improved performance or discarded as unfeasible alternatives, the environmental impact of promising process designs can be assessed. It has been shown that overall, the environmental impact of WWTPs can be decreased through resource recovery implementation [
17]. The growing possibilities of recovery technology integration into treatment processes implies that identifying the most environmentally friendly process alternative requires a careful impact assessment. Life cycle assessment (LCA) is a comprehensive and well-established method to analyze the environmental impact of products, services, and processes. The assessment embraces the entire system involved in the production, use, and disposal of a product or service under investigation. All environmentally relevant substances emitted, as well as extracted natural resources, can be identified and quantified in a “cradle to grave” approach [
109]. LCA allows,, therefore making environmentally beneficial decisions at an early design stage and comparing process designs regarding their impacts. The execution of an LCA should follow a standardized methodology provided by The International Organization for Standardization’s ISO 14000 and 14040 [
110]. A recent review of LCA studies conducted for domestic wastewater treatment plants since the year 1990 concludes that the development of guidelines and standards is necessary to further shape a consistent LCA methodology for the field. For example, different functional units which serve as reference units in LCA are used in different studies which aggravates a comparison of already assessed treatment processes in the environmental dimension [
111].
Since a WRF is in addition to a treatment process and also a production system, the system boundaries of a WRF process LCA will differ from a WWTP LCA, because the recovered and successfully marketed resources avoid the conventional production of similar goods [
112]. The inclusion of these presumable positive impacts is achieved by the so-called approach of consequential LCA. To include the avoided impacts of substituted conventional goods into the assessment, assumptions have to be made on how they are produced. LCA databases provide readily defined impacts for a wide range of different conventional products and materials [
113]. Other needed life cycling inventory data can be collected from published studies in the field of WWTP LCA [
110,
111]. Already available LCAs that include impacts associated with wastewater resource recovery mostly assess both energy recovery [
114,
115] and/or fertilizer recovery [
116,
117] as a consequence of different sludge handling technologies. In addition, the environmental impacts associated with nutrient recovery by aquatic species—like, e.g., algae [
118]—and impacts associated with water reclamation and reuse have been a focus in wastewater treatment-related LCAs [
119].
A conceptual scheme including relevant LCA impact categories of an LCA that assesses the combined impacts of operating a WRF process is drawn in
Figure 3. It has been shown that the impacts of WWTPs mainly occur in the impact categories of (i) eutrophication and (ii) ecotoxicity in effluent receiving water bodies and (iii) global warming potential (GWP) due to sludge handling and electricity use [
110]. In the following, we briefly describe those three impact categories and also which other aspects need to be taken into account when assessing the environmental impact of a WRF process. Since there are a variety of possible negative but also positive environmental impacts, only a complete assessment provides an overview of how an environmentally friendly process design performs in a certain impact category and in total. For more detailed information about the proposed impact categories, we refer to [
120], who provide evaluation criteria and other requirements related to their application in LCA.
The impact category “eutrophication potential” is determined by the effluent concentrations of COD, P, and N that are released into the receiving water body [
49]. They have ideally been estimated by the MEBs during step 3. Heavy metals and micropollutants responsible for ecotoxicity are probably more difficult to estimate due to the early design stage but especially in process designs that apply advanced treatment steps for indirect water reuse, the ecotoxicity impacts can be expected to be significantly lower compared to processes applying only secondary treatment [
121].
The most important emissions for the impact category “global warming potential” relate to the direct GHG emissions of CO
2, CH
4, and N
2O, which can occur at different treatment and resource recovery unit operations. Direct CO
2 emissions from aerobic and biological wastewater treatment are considered by the Intergovernmental Panel on Climate Change (IPCC) of biogenic origin and therefore can be excluded in the GWP estimations of WWTPs [
122]. However, this might not be completely true as [
123] showed that 4–14% of total organic carbon in municipal wastewater is from fossil origin, namely from synthetic products used in industrial and residential products. It is therefore debatable if this minor fraction of the total direct CO
2 emissions should be included in an LCA. WRFs may not only emit CO
2 but also sequester carbon as products recovered from COD, like, e.g., cellulose fibers, may store carbon long term when used as composite construction materials or in other long-lasting applications [
124]. Even waste sludge that is finally disposed in landfills has already been accounted as a carbon sequestration method [
125].
Furthermore, in contrast to direct CO
2 emissions, one has to account for direct CH
4 emissions as, these may represent a significant share of the overall GHG emissions of WWTPs and hence of WRFs. On a time scale of 100 years, CH
4 has a 28 times higher GWP relative to CO
2 [
126], and therefore WRF process designers should be aware of potential direct emission sources and take preventative measures. To assess direct CH
4 emissions, several sources have to be considered. The most severe source (up to three quarters of the total CH
4 emissions of WWTPs) is anaerobic sludge digestion that leads, especially at low temperatures, to a high fraction of CH
4 remaining solubilized in the digestate, from where it can emit to the atmosphere. Those emissions may even exceed emissions avoided through energy recovery from biogas combustion. Therefore, appropriate digestate handling is important like, e.g., the capture of ventilation air applied to sludge handling processes for subsequent use as combustion air in combined heat and power generation [
127]. A critical side stream that contains high amounts of dissolved CH
4 is the supernatant from digestate handling, which is usually recirculated into aerobic treatment units where the CH
4 is biologically oxidized to CO
2 by methanotrophics. This can be improved by different design and/or operation measures; the aeration rate should be high enough to sustain methanotrophic growth in the reactor but low enough to prevent CH
4 stripping. More CH
4 present in the reactor is beneficial for its aerobic conversion compared to low concentrations suggesting to merge CH
4 rich streams into aerobic treatment units; preferably use stirred-tank reactors over plug flow types [
54]. In addition, the influent COD consists approximately of 1% CH
4 that has been produced by microbes in anaerobic zones in the sewer network. This should be taken into account in WRF design because the unit that enters the influent into the WRF should prevent intense contact with the ambient air. For example, screw conveyors imply a more intense contact than centrifugal pumps [
122].
The third GHG that is directly emitted from a WRF is N
2O, which has been estimated to have a 265 times higher GWP than CO
2inn a 100 year time horizon [
126]. The IPCC uses fixed emission factors to estimate N
2O emissions from WWTPs (e.g., 0.035 N
2O-N per kg
−1 influent-TKN), but N
2O emissions depend strongly on the process design. Direct N
2O emissions are, in any case, relevant and hence should be minimized, and may occur at different unit operations that have either anoxic or aerobic zones. In aerobic treatment units, aeration may act as a stripping gas for N
2O [
122]. Although microbial N
2O formation is associated with happening during anaerobic denitrification, more significant N
2O emissions occur during aerobic nitrification due to various species of ammonium oxidizing bacteria (AOBs). Partial nitrification that accumulates NO
2- as an end product is probably the highest potential source that leads to N
2O formation [
122], and therefore a process with integrated nitrification should be designed to prevent that. At low dissolved oxygen (DO) concentrations, the so-called nitrifier denitrification pathway may be used by particular autotrophic nitrifiers that not only oxidize NH
3 to NO
2-, but also reduce NO
2- to NO, N
2O, and N
2, which is optimally only happening in the denitrification process. Nitrifier denitrification contrasts therefore with coupled nitrification–denitrification, where many different organisms oxidize and reduce NH
3 to N
2 step wise [
128]. In contrast, at high DO concentration, the so-called hydroxylamine pathway contributes to N
2O emissions in aerobic treatment units. During nitritation, NH
3 is first converted by AOBs to hydroxylamine (NH
2OH) and then to NO
2-, and N
2O can be formed as a by-product during hydroxylamine oxidation. In addition, partial nitritation is another aerobic process often applied to treat side streams from, e.g., digestate handling, and may lead to N
2O emissions under low DO concentrations [
122].
All in all, these examples show that direct GHG emission prevention by process design is complex and requires a good knowledge of microbial nitrogen and carbon conversion pathways under different operational conditions, like, e.g., differing DO concentrations. It might be difficult to quantify CH4 and N2O emissions accurately at this early design stage, but analyzing the potential critical point sources of a particular process design is necessary to consider emission prevention measures early. For example, choosing those reactor types that allow establishing a homogenous DO concentration in a critical process unit can already be considered at this early design stage to decrease the risk of N2O emissions from aerobic treatment operations.
Another impact category to compare process designs is the amount of hazardous and non-hazardous waste leaving a process. However, we suggest to not include off-site waste handling in the assessment, because data about applied waste management practices of WRF wastes might be difficult to obtain if this is carried out by external companies. When data is available about external waste management practices, it is possible to extend the system boundaries to assess different options like, e.g., the destination of waste sludge (agricultural use, composting, landfill, incineration), which would all have different impacts [
16].
In addition to the described direct environmental impacts of operation, emissions associated with the construction of a process should not be forgotten, as the resource intensity of construction may differ substantially between alternative designs. Those emissions need to be evenly distributed over the time horizon of a planned WRF. In contrast to construction, impacts related to end-of-life phase of a process are usually negligible compared to those from operation and construction [
113]. Additionally, land degradation due to the plant construction might be accounted for since some processes are significantly more compact than others—e.g., aerobic granular sludge treatment would be preferable over conventional activated sludge in this impact category [
129].