A wholelife cost and carbon perspective of alternatives to septic tanks utilising nature-based solutions Open Access

There is growing consensus that the impact of STS on water quality is often understated (Dudley & May 2007) and it is estimated that over 80% of STS in the UK are working inefficiently due to their age (May et al. 2015). Expected effluent quality after the septic tanks for a functioning system is a total suspended solids of 80 mg l⁻¹, biochemical oxygen demand of 90 mg l⁻¹, with the ammonia remaining unchanged from the feed or slightly increasing with a typical level of around 35 mg l⁻¹ (Table 1). Recent findings indicate that in rural communities STS disproportionately affect surface water nitrogen, in comparison to fertiliser usage (Halliday et al. 2014). For instance, ammonia concentrations in ground and surface waters have been shown to more than double downstream of STS (Withers et al. 2012; Herren et al. 2021). In addition, the anaerobic storage of sludge results in release of methane emissions with an estimated rate of 11 gCH₄ PE⁻¹ d⁻¹ equating to 0.11 tCO₂e yr⁻¹ over the 100-year time horizon and 0.34 tCO₂e yr⁻¹ over the 20-year time horizon (European Commission 2022). The regular frequency of desludging also adds further emissions and generates a nuisance impact on local residents. Accordingly, there is a growing need to update the flowsheet for onsite treatment beyond STS to one that improves treatment, reduces greenhouse gas emissions and reduces the frequency of desludging visits. The second ambition aligns to pledges made by the water industry to reach net zero either as full net zero emissions by 2040 (Scottish Water 2020) or operational emissions by 2030 (Water UK 2020a). Beyond the main aspirations there is an appetite to deliver co-benefits to extend the value proposition that any new solution can provide in terms of environmental or societal benefits.

Two alternative approaches are considered (Figure 1). In the first approach, the primary stage is an enhanced septic tank (EST), which combines concepts from septic tanks, anaerobic ponds, and anaerobic baffled reactors (Brown 2023). The system uses baffles to increase digestion thus extending the time between desludging. The effluent from the septic tank is further treated in an aerobic nature-based solution configured as either a passively aerated vertical flow subsurface wetland (VF) or a forced aerated horizontal flow subsurface wetland (AHF) (Figure 1). Both are established technologies used in wastewater treatment and can deliver considerably enhanced effluent quality (Butterworth et al. 2016a). For instance, AHF wetlands have been reported to deliver median effluent qualities of 4 mg l⁻¹ TSS, 4 mg l⁻¹ BOD, and 0.2 mg l⁻¹ NH₄-N when used as a tertiary treatment and 20 mg l⁻¹ TSS, 5 mg l⁻¹ BOD, and 0.8 mg l⁻¹ NH4-N when used as a post primary sedimentation treatment (Butterworth et al. 2016b). Produced methane is collected for reuse or flaring to avoid methane emission into the atmosphere. Recent pilot trials of the NBS post an EST revealed the median effluent of the NBS post EST to be 5 m l⁻¹ TSS, 11 mg l⁻¹ BOD, and 0.5 mg l⁻¹ NH₄-N and 12 m l⁻¹ TSS, 8 mg l⁻¹ BOD, and 3 mg l⁻¹ NH₄-N for the VF and AHF, respectively.

The other option is to use a commercially available package treatment process based on a submerged aerated filter (SAF) to provide aerobic treatment (Figure 1). The system consists of four primary elements: a primary settling tank, an aerated filter, secondary settling tank and sludge holding tank (Figure 1). The sludge from the primary and secondary settling tanks are stored in a sludge holding tank which requires desludging similarly to a septic tank. The certified effluent quality of package treatment SAFs of the style used in this study is 16 mg l⁻¹ TSS, 11 mg l⁻¹ BOD, and 8 mg l⁻¹ NH₄ (Tricel Environmental 2022).

The aim of the current paper is to establish the financial and carbon basis for updating from septic tank systems to alternative flowsheets for population equivalents between 5 and 1,000. To achieve this the four systems were designed at different population equivalents and the wholelife cost (WLC) and lifecycle carbon emissions (LCEs) calculated. To aid in this goal the flowsheets will be considered from two case study perspectives: an old site and a new site. In addition, the paper will also consider the potential societal and ecological impacts of these flowsheets to provide an overall picture of the potential for using EST-NBS-based systems.

The first case study considered an existing site in which an STS is being replaced. The site aims to represent a ‘normal’ site, with dry soil conditions and a soil percolation rate, v_p of 50 mm min⁻¹. The site is situated 40 miles from a large works where desludging tankers and site visits originate from. The site has sufficient topography for gravity feeding of the below ground assets. The second case study considers the construction of a treatment system with equivalent ground and geographical conditions but is a new site. Therefore, the construction of auxiliaries such as security fencing and infrastructure such as roads need to be considered. Both case studies examined 15 population equivalent size flowsheets: 5, 10, 20, 30, 50, 75, 100, 150, 200, 300, 400, 500, 600, 800, and 1,000 PE to assess the impact scale has on the significance of the key design choices. For both case studies, current STS design practice will be considered for a baseline of comparison.

British Water Code of Practice flows and loads (British Water 2013) were used for the influent loading of the flowsheets (Table 1). The design flow of the flowsheets is 0.375 m³ d⁻¹ PE⁻¹ and a peak flow of 0.6 m³ d⁻¹ PE⁻¹ equating to three dry weather flows (DWFs) (Supplementary material, Appendix A). The flowsheets for improved rural treatment works were designed to reach a secondary treatment level (Table 1). Given that STS are unconsented systems (Withers et al. 2012) the STS flowsheets were not designed to treatment consent but instead government recommendation (Scottish Government 2019) to reflect current practice.

Septic tanks in the STS were assumed to have a minimum hydraulic retention time (HRT) of 12 h at peak flow. The tank is installed below ground as per the manufacturer's recommendations. All submerged tanks in this study were assumed to have a maximum design volume of 65 m³. The desludging of the septic tanks is assumed to be yearly based on British guidance (HM Government 2015) and Scottish Water operated tanks (Jefferson 2022). The drainage field is sized according to the Scottish Government guidance (2019) giving an area of 12.5 m² PE⁻¹ for a field with a v_p of 50 mm min⁻¹. The drainage field is assumed to have a maximum length of 20 m (Scottish Government 2019) and be constructed from perforated 13 mm low density polyethylene piping. The pipes are assumed to be installed in 0.8-m-deep and 0.5-m-wide trenches filled with a media of 4–10 mm pea gravel (Scottish Government 2019). The trenches were assumed to be 2 m from any other drainage trenches.

The SAF package treatment plants were selected from available commercial units based on treatment hydraulic loading rates and that certified effluent was below the target range (Premier Tech Aqua 2017; Tricel Environmental 2022). The flowsheets were assumed to be installed as per the manufacturer recommendation (Tricel Environmental 2022). The effluent of the SAF is assumed to be discharged to surface waters.

The EST was designed to have a minimum HRT of 4 h at peak flow, with four internal baffles. This was based on recent low temperature hydrolysis characterisation and pilot-scale reactor testing, which showed baffles were required to produce an extended desludging interval of 7 years rather than simply increasing the overall septic tank volume (Jefferson 2022). EST systems were assumed to be constructed through earthworks, with a lined pit and soft cover made from synthetic rubber, using concrete posts for baffling. The wetlands were designed to current standard design guidance (Dotro et al. 2017) to mirror current practice in the UK (Butterworth et al. 2016b). The VF wetland area was determined to ensure they had a peak flow hydraulic loading of 0.12 m³ m⁻² d⁻¹ with a square area and with a treatment depth of 0.8 m and freeboard of 0.3 m, the maximum length of a single VF wetland is 25 m. The media comprised of a main treatment layer of 55 cm of sharp sand and 10 mm shingle mixed at a 2:1 ratio with the final 5 cm layer of 10 mm shingle. The treatment layer sits above a drainage layer of 20–40 mm gravel for the bottom 15 cm and a 5-cm transition layer of 10 mm shingle (Jenkins 2017). Distribution and collection pipes, constructed from PVC piping, are positioned every 1.25 m on the surface and bed of the wetland with 0.5 × 0.5 m concrete splash plates placed every 1.25 m (Jenkins 2017). The system is dosed 8 times a day with 5 minutes doses delivered by a bell siphon within the EST.

The AHF wetlands design is based on a standard horizontal subsurface flow wetland used for tertiary treatment in the UK but with the adaptation of having a coarse bubble aeration grid installed on the flow of the wetland as is current practice (Butterworth et al. 2016b). Accordingly, the wetland was sized with a maximum treatment width of 25 m and organic loading below 15 g_BOD m⁻² d⁻¹ with an air flow of 0.26 m³ PE⁻¹ h⁻¹ (Butterworth et al. 2016b). The AHF was assumed to have a treatment height of 0.6 m and a freeboard of 0.3 m containing 4–10 mm pea gravel as the main treatment media and distribution layers of 100 mm stone for the first 0.5 m from the inlet and outlet, held in place with steel gabions.

Fences were assumed to be 1.8 m steel palisade fence posts with a 3 m wide access gate (Water UK 2020b). Road infrastructure design was outsourced to quoting contractors consisting of two road designs, gravel and bitumen. Gravel roads were used when desludging visits were less than every 3 years (Jefferson 2022) otherwise bitumen roads were assumed. The drainage fields of STS are fenced off as per Scottish Water guidance.

STSs are not monitored beyond desludging. In contrast, the alternative flowsheets were assumed to be monitored monthly with electrical components inspected yearly. Replacement frequencies were assumed to be yearly for the filters, every 3 years for silencers, 5 years for bearings, and 15 years for non-return and pressure relief valves. The operator time is contracted as £35 h⁻¹ with inspections lasting 1 h. Desludging was assumed to be conducted by diesel tankers, with the volume of sludge removed assumed to be 33% of the total tank volume for STS, ESTs, and SAFs. Inspection and maintenance vehicles are modelled as diesel vans with travel originating from a large works site.

Costs were calculated at June 2021 values, with costs adjusted by the Consumer Price Index (CPI) (ONS 2022). Flowsheets were calculated through component costing, with each element worth more than £50 quoted in triplicate. Where off the shelf products were not available cost curves were used to determine the value of the component. The system lifetime of the project was assumed to be 30 years (Gallagher & Gill 2021).

Flowsheet process emissions are derived by best available data in the literature (Supplementary material, Table A1) and biogas flaring was modelled as flaring of sour waste natural gas with biogenic carbon dioxide release from the Ecoinvent v3.8 database. Desluding tankers were modelled as unladen on the outward journey and the additional sludge weight is assumed to have a density of 1,000 kg m⁻³. Desludging tankers were modelled as the average carbon emissions per tonne mile of Euro 3 through to six lorries. Three lorry weight classes are considered, 3.5–7.5, 7.5–16, and 16–32 tonnes assuming capacities of 2, 10.5, and 19 m³, respectively. The carbon emissions of UK grid electricity up to 2040 are based on UK department for Business, Energy and Industrial strategy projections (2019). Post 2040, electrical grid emissions are based on sum of root square error with an exponential decay model, modelled using MATLAB (Supplementary material, Appendix B).

Uncertainty of the capital emissions was calculated using a Monte Carlo method with 1,000 iterations (Heijungs 2020), using the uncertainty within the Ecoinvent database. Due to the large variability in reported process emissions for each system, the uncertainty in process emissions was considered qualitatively.

Comparison across the different alternative options shows that the EST-NBS flowsheets are a lower cost upgrade to STS than SAFs (Figure 3). The WLC per PE decreased for all options but at different rates such that the WLC per PE for the EST-AHF and the STS reached parity at sizes of 150 PE and above. The critical components that delivered the cost saving per PE in the EST-AHF were the operating costs of the AHF compared to the operating components in the STS as the capital cost per capita are approximately equal. Operating costs were 81% higher than those for the STS at small scales (Figure 3) but the difference decreased with increasing scale such that at 1,000 PE the operating costs of the STS and EST-AHF are within 4%. The principle operating cost of the STS is desludging, which decreases from £776 to £226 PE⁻¹ as the scale increases from 10 to 1,000 PE. In comparison, the main operating costs for the EST-AHF are the operating and maintenance costs associated with the AHF which decreases from £1,350 to £172 PE⁻¹ from 10 to 1,000 PE.

Similarly, the WLC per PE of SAFs and EST-VFs is comparable beyond 150 PE but higher than the STS. For instance, at 1,000 PE the SAF and EST-VF represent an additional 70 and 54% WLC compared to the STS, respectively. However, sub-150 PE, the EST-NBS flowsheets represent a notable saving compared to SAFs and hence offer a lower cost option for enhancing treatment effectiveness compared to an STS. To illustrate, at 100 PE the WLC of an EST-VF and EST-AHF are £786 and £613 PE⁻¹ compared to £1,080 PE⁻¹ for a SAF. While the EST-VF is a more expensive option than the EST-AHF at larger scales, below 30 PE the EST-VF is the least cost option for improved treatment. This reflects the low operating cost of the VF system. At 10 PE the VF has a lifetime operating cost 25% lower than the AHF, however as the economies of scale reduce the impact of fixed factors such as operator transport and time, the increased capital cost of the VF system means that the EST-AHF is a lower cost alternative at larger scales.

Analysis of the cost breakdown reveals that the operating costs dominate the WLC at smaller scales but become a more minor component as scale increases reaching approximate parity with CAPEX at a scale of 30 PE, 150 PE, 75 and 1,000 PE for the SAF, STS, EST-VF, and EST-AHF, respectively. To illustrate the contribution of operating costs, at a scale of 10 PE, was 79% for the SAF, 71% for the STS, 81% for the EST-AHF, and 66% for the EST-VF (Figure 3). In comparison, at 1,000 PE the OPEX contributes 30% of the WLC per PE for the SAF and 31% for the EST-VF.

The change in WLC per PE can be utilised to derive an economy of scale exponent assuming a standard power law fit. The different options group into two sets with the exponents of 0.61 and 0.66 for the SAF and EST-AHF and 0.80 and 0.82 for the STS and EST-VF. This compares to a to standard value of 0.66 reflecting the dominance of civil structures on the overall costs for the SAF and EST-AHF. Whereas the higher values of the STS and EST-VF indicate towards more modular components similar to the exponents reported for membrane based systems (Jefferson 2022).

The LCE for the STS has a negligible scale impact ranging from 4,260 kg CO_2eq PE⁻¹ at small scale to 4,210 kg CO_2eq PE⁻¹ at larger scales due to the fact that the majority of the LCE is derived from process emissions of methane. This is in line with other studies into the lifetime carbon footprint of septic tank flowsheets which indicated that process emissions account for 90% of the total LCE (Resende et al. 2019) compared to 85% in the current study. While the STS directly vents methane into the atmosphere, the other options either prevent its formation by utilising aerobic conditions (SAF) or capture it (EST-VF; EST-AHF). The impact is to reduce the LCE associated with process emissions from 129 kg CO_2eq PE⁻¹ yr⁻¹ for a 100 PE in the case of an STS to 0.98, 0.88, and 0.80 kg CO_2eq PE⁻¹ yr⁻¹ for SAF, EST-VF and EST-AHF at the same scale. Over the 30-year lifetime of the assessment, the capture of methane in the EST-NBS systems avoids 3.8 t CO_2eq PE⁻¹ compared to the STS, this is the equivalent to 3 flights from London to New York (Ecoinvent database 2022).

With respect to a centralised wastewater treatment, consisting of an activated sludge system and anaerobic digester, the operational emissions per capita are approximately 82 kg CO_2eq PE⁻¹ yr⁻¹ (Piao et al. 2016), which is equivalent to a SAF flowsheet and greater than either EST-NBS flowsheet which are between 10 and 54 kg CO_2eq PE⁻¹ yr⁻¹. The operational emissions of an STS system is greater than that of a centralised works on a per capita basis.

The uncertainty in the capital emission of the LCE is low (Figure 4) with no significant impact on findings. On the other hand, the determination of process emissions remains the greatest area of uncertainty in the estimation of the LCE due to a lack of reported emission rates and the overall consistency of approach. To illustrate, in the case of the SAF a value of 64 gCO_2−eq m⁻³ wastewater treated was derived from nitrous oxide emissions for SAFs from UK carbon accounting guidance (UK WIR 2008) and methane emissions from sludge holding tanks of IPCC (IPCC 2014). However, in a recent study from China, emissions ranged from 264 to 443 gCO_2eq m⁻³ wastewater treated with an average of 333 gCO_2eq m⁻³ wastewater treated (Hua et al. 2022). The consequence is a potential increase in process emission of five times, based on average and nine times, based on the maximum level reported which results in an additional 4 kg CO_2eq PE⁻¹ yr⁻¹, however, this increase does not significantly impact the findings of this study in terms of technology comparison (Figure 5).

In the case of the AHF, there is no reported data on emissions post an anaerobic process. The closest approximated reported case is for treatment of fish farm effluent with an emission rate of 1.62 kg CO_2eq PE⁻¹ yr⁻¹ (Maltais-Landry et al. 2009), the methane flux is in line with that reported for domestic AHF systems (Wang et al. 2008). Reported data does exist for the vertical flow systems with projected process emissions ranging from 0.16 to 4.6 kg CO_2eq PE⁻¹ yr⁻¹ (Mander et al. 2014). Nitrous oxide accounts for 10% of the process emissions of VF wetlands and 100% of AHF wetlands. This compares to 83% for the SAF revealing that management of the nitrogen treatment pathway is critical in minimising process emissions from such flowsheets. The general indications of how to minimise nitrous oxide emission are linked to avoiding overloading, running as close to steady state as possible and ensuring sufficient oxygen is distributed through the biofilm (Kampschreur et al. 2009). Comparing the operation of the NBS wetlands, the AHF runs in a continuous mode opposed to the batch nature of VFs. However, despite the expectation of an AHF having a lower N₂O emission rate than a VF, currently available literature gives a nitrous oxide emission rate for an AHF as 1.12 g N₂O kg TN_treated⁻¹ (Maltais-Landry et al. 2009) compared to 0.18 g N₂O kg TN_treated⁻¹ for a VF (Mander et al. 2014). Measurement and establishing conditions to minimise overall process emissions remains an area of need and a recommendation for future work.

Interestingly, the estimation of methane emissions from septic tanks is consistent across different studies, with numerous studies reporting a value of 11 gCH₄ PE⁻¹ d⁻¹ (Diaz-Valbuena et al. 2011; Truhlar et al. 2016; Huynh et al. 2021). This equates to an overall emission rate of 112 kg CO_2eq PE⁻¹ yr⁻¹.

The other area of major uncertainty is related to the desludging frequency and distances travelled. The desludging frequency of the packaged treatment systems and the septic tanks is approximately 1 year and this compares to the EST-NBS systems of 7 years. However, the desludging frequency of water utility operated septic tanks is known to vary from 3 to 24 months and is commonly accomplished through an integrated schedule to desludge multiple septic tanks in individual tankering trips. The contribution desludging makes to the operational emissions is 1, 8, 1, and 1% at 100 PE with a distance of 40 miles for the STS, SAF, EST-VF, and EST-AHF, respectively. If both the distance travelled and desludging frequency were double that of the modelled systems (or equivalent) there would be a 13, 30, and 19% increase in the LCE of the STS, SAF, and EST-NBS flowsheets, respectively. This could be the case for particularly remote sites such as flowsheets situated on small islands where tankers are required to travel back to the mainland. However, although the desludging can have a significant impact on the LCE it does not impact the findings with respect to flowsheets with reasonable operational variation (Figure 5(b)). For example, with a travelling distance of 40 miles per trip it would require desludging every 60 days for the EST-NBS flowsheets to have an equivalent LCE to a SAF.

Over the predicted life span of 30 years, the relative cost of upgrading to an EST-VF at 10 PE is £560 PE⁻¹ which produces a lifetime carbon reduction of 2,520 kg CO_2eq PE⁻¹, giving an abatement cost of £220 tCO_2eq⁻¹ over 30 years. Comparatively at 10 PE, the abatement cost of a SAF and EST-AHF is £1,860 and £260 tCO_2eq⁻¹. At larger populations, the abatement costs of the alternative options are reduced as the price differential from an improved flowsheet to STS decreases while the carbon reduction increases. For example, at the 1,000 PE scale, the abatement cost of a SAF, EST-VF and EST-AHF are £92, and £65, and –£17 tCO₂ eq⁻¹. The negative number represents a cost-effective investment, providing an operational saving alongside carbon reduction, for example, a reduction in travel both saves money and fuel and the associated carbon emissions. The significances of the estimated abatement costs can be considered by framing the number within the general abatement costs published for selected technologies for the water sector (Table 2). Overall, this demonstrates that the proposed alternatives can provide a cost-effective means of carbon abatement while delivering a higher level of treatment. To illustrate, at 1,000 PE an EST-AHF has a lower lifetime cost and emissions than STS meaning the abatement cost is –£17 tCO₂ eq⁻¹, a more cost-effective abatement technique than the injection of biomethane to the grid.

In the case of new sites, additional consideration needs to be given for infrastructure costs associated to road and fencing. For instance, if the site requires regular desludging visits such as with the STS and SAF, a bitumen road is required (Jefferson 2022). If the desludging frequency is less than every 3 years, then gravel road is considered suitable (Jefferson 2022). The impact is seen through the difference in cost and emissions of the two options. Bitumen based roads have a cost of £130 m⁻¹ and emissions of 66 gCO_2eq m⁻¹ lane⁻¹ compared to £26 m⁻¹ and emissions of 4.9 gCO_2eq m⁻¹ lane⁻¹ for gravel-based roads (Espinoza et al. 2019). Consequently, there is a differential cost of £104 m⁻¹ and an emission of 61.1 gCO_2eq m⁻¹ lane⁻¹. The exact impact will depend on the specific distances associated with a given site but applies for all options. Consequently, the cost and emissions for the NBS system will always be lower, further reinforcing the positive impact compared to a SAF or STS observed for the existing site case study.

The other major additional cost is associated with fencing which is dependent on the land footprint of the system. The footprint of the different options at 100 PE was 1,260, 50, 589, and 360 m² PE⁻¹ for the STS, SAF, EST-VF and EST-AHF, respectively. This converts to a fencing cost of £10,000 for STSs and £2,600, £4,100, and £6,000 for SAF, EST-VF and EST-AHF, respectively. The lower costs for the EST-NBS over the STS represent the cost associated with the drainage field. This equates to a 22% increase in the lifetime costs of an STS, however for the improved systems the impact is lower being 3, 8, and 7% for SAF, EST-VF, and EST-AHF, respectively. Overall, fencing does not significantly change the assessment of costing compared to an existing system. For instance, at 100 PE an STS remains the lowest cost system, followed EST-AHF, then EST-VF and finally SAF.

In more rural sites there is the possibility that an electrical connection is not possible. The SAF and EST-AHF are dependent on an electrical connection to operate, compared to an STS and EST-VF which can operate passively. The cost of installing mains electricity connection is £13,000 for a small business and can rise to up to £50,000 (Northern Power Grid 2022). Onsite power generation could also be used but it is likely that a redundancy connection would be required. Therefore, in these scenarios, the adoption of a passive system such as the EST-VF would likely be beneficial.

A key to the lower LCE for the EST-NBS options is the fact that no methane is vented into the atmosphere. However, if the gas from the EST systems were vented then the overall LCE would be unfavourable compared to the STS and show the importance of gas management. The retrofitting of flares to STS could provide a low-cost method of reducing the carbon load of STS, however it will not provide the additional treatment capabilities of the improved flowsheets and therefore is not a suitable solution if increased treatment is required.

Instead of flaring, the produced methane could be utilised to provide heating for the EST or used within the local community providing the potential for further carbon offsets. A similar proposal for a sewer mining project to irrigate a park in Australia proposed the use of the gas produced to power several BBQ rings to enable people visiting the park to cook their lunch (Jefferson 2022). While there is currently insufficient knowledge to assess the feasibility of the different options, it provides an illustration of approaches to add additional value from the EST-NBS flowsheets.

Nature-based solutions to rural wastewater provide a lower cost method of improving treatment of septic tank systems compared to SAFs at both new and existing sites. In addition to being a cost-effective solution, the NBS flowsheets have lower carbon emissions than more intensified SAF at populations from 5 to 1,000 PE. The work shows the potential that additional focus on rural treatment can have on the carbon emissions of the UK water sector, comparable in cost effectiveness to upgrading large anaerobic digesters and biomethane grid injection. This is coupled with improved water treatment, reducing the eutrophication burden which STS impose. There is scope to expand the flowsheet to provide a social space and societal gain, alongside treatment and carbon reduction.

The authors gratefully thank Scottish Water for part-funding and technical support for this project. The financial and practical support of EPSRC is also appreciatively acknowledged. Data underlying this paper can be accessed at DOI: 10.17862/cranfield.rd.21711260.

All relevant data are available from an online repository or repositories (https://cord.cranfield.ac.uk/articles/software/Data_underpinning_NERC_Research_Translation_Grassland_Management_project/11359613).

The authors declare there is no conflict.