System Boundary & Functional Unit
The two systems being compared are consistent with those included in the cost analysis section of this report. Table 15 provides a summary of the steps modelled for each of the two cleaning regimes. Important considerations are that:
• Given the likely variation in actual cleaning regimes employed at wastewater recycling facilities around the world, the ‘conventional’ scenario represents a hypothetical cleaning regime in order to provide indicative results.
• The FNA cleaning regime is based on a number of important assumptions that could vary depending on actual plant circumstances.
Table 15. The 'Conventional' and 'FNA' cleaning regimes considered in the life-cycle analysis.
Cleaning Step ‘Conventional’ scenario ‘FNA’ scenario
Initial membrane flush
Low pressure flush with RO permeate for 10min.
All flush water discharged to a sump, then pumped to the head of the STP.
Low pressure flush with RO permeate for 10min.
All flush water discharged to a sump, then pumped to the head of the STP. Prepare the primary
cleaning agent
Pump 50%ww caustic solution from bulk storage tank into the chemicals mixing tank,
and dilute to 0.1%ww with RO permeate. Heat the solution to 35°C.
Mix solid NaNO2 & 33%ww liquid HCl with RO permeate in the mixing tank, to achieve
a nitrite concentration of 50 mg/L at pH 3. Circulate the primary
cleaning agent
Circulate the solution through the membranes for 60min, and then soak.
Circulate the solution through the membranes for 120min, and then soak. Flush the cleaning
agent out of the membranes
Low pressure flush with RO permeate for 10min.
All flush water discharged to a sump, then pumped to the head of the STP.
--
Prepare the secondary cleaning agent
Dilute 33% liquid HCl with RO permeate in the chemicals mixing tank, to achieve a 0.2%
HCl solution at pH 1-2. Heat the solution to 35°C.
-- Circulate the
secondary cleaning agent
Circulate the solution through the membranes
for 60min, and then soak. --
Final membrane flush
High pressure flush with RO permeate for 10min.
Low pressure flush with RO permeate for 60min.
All flush water discharged to a sump, then pumped to the head of the STP.
High pressure flush with RO permeate for 10min.
Low pressure flush with RO permeate for 60min.
All flush water discharged to a sump, then pumped to the head of the STP.
System boundary
background database info onsite operations RO membrane unit process water concentrate chemicals mixing tank permeate chemicals manufacture transport to site membrane manufacture transport to site transport spent membranes from site disposal spent chemicals to sewer material &energy inputs emissions to air,water & land
discrete scenarios compared significance of potential variation to be considered through sensitivity analysis
electricity generation electricity used on site permeate storage tank RO pre- treatment
Figure 35. The scope of the system flows that were included in the analysis for each of the two scenarios.
Key elements that define the system boundary are:
• Only operational flows were included in this analysis. Construction of the treatment plant infrastructure, and end-of-life disposal of that infrastructure, were excluded on the grounds that (a) they are typically only a small contributor to the overall life-cycle impacts associated with urban water systems [see 77, 78, 79]; and (b) the requirement for infrastructure (e.g. tanks & pumps) is expected to be largely the same for the two different cleaning regimes.
• The life-cycle inventory includes the generation & supply of the chemicals used for the cleaning regime. While the discharge of spent cleaning fluids is included in the modelling, any downstream implications of treating that waste are not considered in this analysis.
• The generation of RO permeate is also included as an input to the system, incorporating a full life-cycle model of the RO permeate production system. This includes all material and energy inputs to the production plant, along with the disposal of MF backwash and RO concentrate flows. The model used here excludes any operations downstream of the RO membrane (e.g. advanced oxidation, disinfection & product water discharge pumping).
• This recognises that if less permeate is used for the cleaning process (flushing & mixing with chemicals), then that permeate would become available as product water that can be delivered to users.
• Reducing the permeate requirements for membrane cleaning would therefore reduce the overall production required by the treatment plant, and the environmental implications of that reduced production are attributed to the scenarios shown here.
The basis for analysing and comparing the two scenarios (the functional unit) is a single cleaning cycle. Implicit in this choice of functional unit is the assumption that the FNA-based cleaning regime would be implemented with the same frequency as the more conventional alternative. The implications of that assumption are considered further below.
Results are presented on a per-membrane or per-plant basis depending on the question being addressed in each section of this chapter - in all cases; it was assumed that the input flows are proportional to the number of membranes included.
Inventory data
For each of the system flows described in Table 15, mass and energy flows were estimated using equipment specifications taken from an existing facility for cleaning RO membranes in an advanced wastewater treatment plant.
Database and literature sources were used to estimate the material inputs, energy inputs, and environmental emissions associated with providing the inputs (chemicals; RO permeate; membranes; electricity) to the cleaning process (Table 17).
Impact Assessment
The analysis in this chapter focusses primarily on greenhouse gas (GHG) footprints, using 100yr equivalency factors taken from the most recent IPCC report [80].
Twelve other life-cycle impact assessment (LCIA) categories were also considered briefly in the final part of the results section below. The indicators are described in Table SI 12 (Appendix K). The suite of categories, and the models chosen for each, followed the default approach chosen in [77].
Software
The inventory models, and impact assessment results, were implemented in the Simapro (v8.05) software.
Table 16. Estimates of the foreground inventory flows for each of the scenarios.
Inventory item Conventional scenario
FNA scenario
Initial membrane flush
336 L of permeate used per membrane being cleaned (0.56 L/s per membrane, for 10min)
Prepare the primary cleaning agent
0.2 kg of 50%ww caustic per membrane
112 L of RO permeate per membrane 1.3kWh of electricity per membrane
0.03 kg of solid NaNO2 per membrane 0.05 kg of 33%ww HCl per membrane 111 L of RO permeate per membrane Flush the
cleaning agent out of the membranes
336 L of permeate used per membrane being cleaned
(0.56 L/s per membrane, for 10min) -- Prepare the
secondary cleaning agent
0.7 kg of 33%ww HCl per membrane 111 L of RO permeate per membrane
1.3kWh of electricity per membrane
-- Final membrane
flush
2750 L of permeate used per membrane being cleaned
(1.22 L/s per membrane for 10min; then 0.56 L/s per membrane for 60min) Bulk chemical
transfer pump 60 W motor operating at 1.5 L/h
Permeate supply
pump 18.5 kW motor operating at 22 kL/s
Chemical
circulation pump 1.2 kW power required per membrane being cleaned
Sump pump 7.5 kW motor operating at 35 L/s
Table 17. Sources & assumptions used for the background inventory estimates.
Inventory item Source Key assumptions
Manufacturing & transport of chemicals used in the cleaning process
Australasian LCI database [81]
All chemicals are manufactured in Australia. The manufacturing inventories are based on the Ecovinvent v3 database [82], modified to use Australian inputs (e.g. power & transport
models) Generation & supply of
electricity used in the cleaning process
Australasian LCI database [81]
Electricity is assumed to be from a system average mix of Queensland generators Production of the RO permeate Data from Seawater desalination & Advanced Wastewater
Treatment Plant to recycle wastewater for Direct Non-potable Reuse [77]