Life Cycle Assessment of Stormwater Treatment Systems

Life Cycle Assessment of Stormwater Treatment Systems R Bennett and D Bishop Final Year Projects, 2011 Dept. of Civil and Natural Resources Engineerin...
Author: Neil Powell
1 downloads 0 Views 849KB Size
Life Cycle Assessment of Stormwater Treatment Systems R Bennett and D Bishop Final Year Projects, 2011 Dept. of Civil and Natural Resources Engineering University of Canterbury Project supervisor(s): A O‟Sullivan and D Wicke Keywords: Rain Garden, Ecological Engineering, Hynds Downstream Defender ABSTRACT Stormwater treatment is becoming more prominent in practice due to increased public awareness and more stringent guidelines in compliance to the Resource Management Act 1991. Various treatment options are employed, including proprietary systems such as the „Hynds Downstream Defender‟, which operates on a vortex hydraulic head mechanism and ecologically engineered bio infiltration systems such as „rain gardens‟. Stormwater treatment systems are sized to treat one third of the volume of water from a storm with a return period of 2 years and duration of 24 hours. Irrespective of their type, treatment systems in New Zealand are designed to remove at least 75% of the total suspended sediment load and it is assumed trace metals, nutrients and oil are concurrently removed. While stormwater treatment is inherently better than no treatment there is limited information about the environmental impacts of the choice of stormwater treatment system. The environmental impacts associated with the raw materials, construction and operation and maintenance of the treatment system are quantifiable through a life cycle assessment (LCA). LCA‟s provide an indication of the true sustainability of the stormwater treatment system through a „cradle to grave‟ assessment approach. Financial aspects are not incorporated into LCA analysis, but it is often to perform a cost benefit analysis. In this study life cycle assessments were carried out for a proprietary concrete detention unit (Downstream Defender), a subsurface sand filter and a conventional vegetated rain garden (designed to New Zealand guidelines). SimaPro 7.3 LCA modelling software was used, with life cycle inventory information gathered from the ecoinvent database and impact assessment carried out using ReCiPe 2008 following international ISO 14040 protocols. Key environmental impact metrics included in the assessment were climate change (kg CO2 equivalence), terrestrial acidification (kg SO2 equivalence), freshwater (kg P equivalence) and marine eutrophication (kg N equivalence) and terrestrial Eco toxicity (kg 1,4 dichlorobenzene equivalence). Results show that the vegetated rain garden design has the largest impact across all categories followed closely by the Downstream Defender. The system with the overall lowest impact is the subsurface sand filter. The magnitude of the impact from climate change was determined to be significantly larger than all other categories considered. Compost used in the vegetated rain garden was shown to have a significant contribution to the process of „raw materials and construction‟, while road transport was shown to make a large contribution to environmental impacts, especially in regards to climate change. LCA provides a value tool to inform stormwater management decisions.

1.

including „rain gardens‟ (Hynds Environmental, 2006) (ARC, 2003).

INTRODUCTION

The importance of urban stormwater treatment is appreciating. Stormwater management and design guidelines have evolved over time, with improved understanding of the downstream effects of untreated stormwater on natural ecosystems. Urban stormwater in New Zealand typically contains high levels of heavy metals, such as zinc, copper and lead, contributing to this is the wear-and-tear of vehicle parts (Sansalone and Buchberger, 1997). In 2003, a paradigm shift occurred in Christchurch, in that stormwater treatment moved away from the primary concern of drainage, to a more holistic approach incorporating six values; culture, heritage, ecology, recreation, landscape and drainage (CCC, 2003). Various forms of stormwater treatment are prevalent in Christchurch, from proprietary „off the shelf‟ systems such as the „Hynds Downstream Defender‟ to more ecologically designed systems

The treatment performance and design of stormwater best management practices (BMPs) are defined in guidelines, such as the Auckland Region Council‟s Technical Publication 10 (ARC TP10). In many situations BMPs are employed with little regard to their net environmental impact. Direct impacts of construction and indirect impacts of the embodied energy and materials of the system are typically ignored (Kirk et al, 2006). Environmental impacts associated with the implementation of a stormwater treatment system are quantifiable using a life cycle assessment (LCA) methodology. The life cycle typically considers from „cradle to grave,‟ this includes the extraction and processing of raw materials, manufacturing of the

1

system, transportation of materials and equipment to the site, construction and operation of the system, maintenance and the final decommissioning of the system (Wimmer et al, 2004). No financial aspects are incorporated in the LCA but in practise a cost benefit analysis would be considered in parallel.

cycle assessment (LCA). Data for the materials and processes in the lifecycle were sourced from ecoinvent database produced by the Swiss centre for life cycle inventories (Frischknecht et al, 2007). Impact assessment was carried out using ReCiPe 2008 using the impact categories of climate change, terrestrial acidification, freshwater eutrophication, marine entrophication and terrestrial ecotoxicitiy (Goedkoop et al, 2009). These categories are discussed below.

Life cycle assessment methodology has numerous applications in determining the long term, indirect and cumulative impacts of human actions and to date has been used in regard to building design and electronic equipment (Kirk et al, 2006). LCA studies for stormwater systems are not documented in literature. In a New Zealand study a life cycle approach is used to compare the energy requirements and carbon dioxide emissions associated with a rain garden and sand filter but LCA modelling software was not utilized (Andrew and Vesely, 2008). Results from this study showed that over a life-time of 50 years the rain garden had 20% less energy requirements and 30% CO2 emissions. Transportation was shown to have a significant impact upon the results.

The processes modelled for each treatment system were the extraction and processing of raw materials, manufacturing the system, transportation of materials and equipment, excavation and construction of the system. The boundary for this analysis is shown in figure 1. Pre-treatment elements, such as the gross pollutant trap which remove leaves and other debris from the stormwater, were not included in the analysis.

In this study three stormwater treatments were compared using a life cycle assessment to better understand their environmental costs. The three systems selected were a subsurface sand filter, a proprietary concrete detention unit and a conventional vegetated rain garden. The fundamental hypothesis of this research is that an ecologically engineered stormwater system has a lower environmental impact than a proprietary system. This can be quantified through ISOapproved LCA methodology.

Figure 1. System Boundaries of Treatment Systems used in model. Operation and Maintenance was not included in the final model within SimaPro as ecoinvent was not able to provide the necessary information. 2.2. Inventory analysis

2.

METHODS The three systems selected were a subsurface sand filter (located in Addington, Christchurch), a proprietary concrete detention unit („Hynds Downstream Defender‟) and a conventional vegetated rain garden (hypothetical design from NZ technical guidelines).

LCA modelling was conducted using the programme SimaPro 7.3 (PRé Consultants, Netherlands) following ISO 14040 protocols (2006). There are four distinct phases in the life cycle assessment; goal and scope definition, inventory analysis, impact assessment and interpretation of results (table 1).

The subsurface sand filter model is an as built design in Christchurch (shown in figure 2). The catchment for this system was used for the design of the other two systems to ensure consistency in runoff volumes to be treated. The residential area draining into the system is 4020 m2 and contains 65% impervious surfaces (Good, 2011).

Table 1. Life Cycle Assessment Framework (Kirk et al, 2006) Goal and Scope Definition Life Cycle Inventory Life Cycle Impact Assessment Interpretation

Establish objectives, determine methods, define system boundaries Catalogue materials and processes Determine environmental consequences Draw conclusions, assess sensitivity

In New Zealand rain gardens are designed using Auckland Regional Council (ARC) Technical Publication 10. Chapter 7 of this guideline is based on filtration design, construction and maintenance. Although the Addington treatment system is called a „rain garden‟ by the Christchurch City Council it is more strictly a subsurface sand filter according to ARC guidelines.

2.1. Goal and Scope

To contrast with the as-built Addington system a hypothetical rain garden was designed to ARC Technical Publication 10 guidelines (see figure 3). It has

In this study the environmental impacts of three stormwater treatment system were compared using a life

2

The ecoinvent database was used to determine impacts from extracting and processing raw materials and the emissions associated with transportation and excavation for each stormwater treatment system. The data entered into the ecoinvent database is of high quality and includes areas of energy, bioenergy, transportation, waste management, construction, chemicals, and agriculture (Frischknecht et al, 2007). The ecoinvent database was created in the year 2000 and is predominantly based on European conditions. This database is one of several that are included within the SimaPro interface. Ecoinvent database was used as it contains the most complete range of information needed for this LCA.

Figure 2. Addington stormwater system schematic (note: not to scale) (Good, 2011)

For each stormwater treatment system the processing of raw materials, transportation to the site and excavation were considered, as shown in figure 1. When transportation of materials was not known an estimated travel distance was assumed. Operation and maintenance, although an important consideration throughout the life cycle of a stormwater treatment system, were not included in the final model. Sufficient information was not available in the ecoinvent database to accurately portray the process. Additionally, with a stormwater treatment system the maintenance regime is often finalised after the installation of the system. Maintenance is typically site-specific and timing of inspections are determined a period of time after the system has been operating.

Figure 3. Schematic of Rain garden, designed to ARC Technical publications 10 guideline (note: not to scale)

2.2.1 Raw Materials and Construction Raw materials used in each system and the source location of these materials are summarized in table 2. The Addington system makes use of a perforated plastic „Versitank‟ which detains stormwater and allows for fine particles to settle out before the water is discharged to groundwater. The materials of the rain garden designed according to the ARC TP10 were stipulated by the guideline. The rain garden was modelled as a system that discharges to surface water. To restrict discharge to ground water a geosynthetic clay liner (GCL) is integrated into the design. The GCL incorporates a clay layer of low hydraulic conductivity, typically bentonite.

Figure 4. Schematic of Hynds Downstream Defender (note: not to scale) (Hynds Environmental, 2006) been assumed that the TP10 guidelines are appropriate for Christchurch conditions, although were produced for Auckland. The system was entitled „TP10 rain garden‟ and was designed for the same catchment as the Addington system. The rain garden was sized for a storm with a return period of 2 years and duration of 24 hours. This was determined to be 53 mm for the given catchment using available information from NIWA (2009). The surface area for the rain garden was determined to be 55 m2.

Information about raw materials utilized in the design of the Hynds Downstream Defender (table 2) were sourced from Hynds Environmental (Amat, pers. comm, 2011). The system is manufactured in Auckland, primarily from materials sourced internationally. 2.2.2 Transportation Transportation was incorporated into the model by defining the distance from the source of raw materials to the final site at Addington, including manufacturing/processing sites along route. Transportation within the model is measured in units of tkm where 1 tkm represents 1 tonne transported a distance of 1 km. In the case of the TP10 rain garden

The proprietary concrete detention unit was modelled as a „Hynds Downstream Defender‟ (see figure 4). A range of model sizes are available and the selection is directly based on the design flow rate. For the Addington catchment the design flow rate was determined to be 11.7 L/s. The 1200 diameter unit is therefore most appropriate in the Addington catchment.

3

TP10 Rain Garden Design

Hynds Downstream Defender

Addington

Table 2. Modelled stormwater treatment system materials and source locations Material Gravel Swale A2 Sand Versitank Weedmat Pipe Concrete

Used in Model Gravel Sand PVC plastic Polypropylene PVC plastic Concrete Block

Mass 12,700 kg 2,100 kg 72 kg 58 kg 2 kg 3,130kg

Cast Iron Polyethylene Internal Components Galvanised Steel Gravel under-drain Planting soil Sand Ground cover Geosynthetic clay liner Pipe for underdrain

Cast Iron HDPE

70 kg 40 kg

Low alloyed steel Gravel Compost Sand Bark Bentonite clay PVC plastic

25 kg 37,000 kg 33,000 kg 26,400 kg 660 kg 400 kg 3 kg

design, where material locations were not predefined, materials were sourced from the closest possible location, preferably within the Canterbury region. For example, gravel used in both the Addington system and TP rain garden design was sourced from the Waimakariri River. The material was then transported to a depot in Hornby for sorting and sizing. The material was then transported to the Addington site for construction of the system.

Source Location (distance to site) Waimakariri River (25 km) New Brighton (30 km) Sydney, Australia (2,600 km) Sydney, Australia (2,600 km) North Canterbury (47 km) Cement from Sydney Australia, Concrete produced in Auckland (2,200 km to Auckland) China (10,300 km) Sydney, Australia (2,200 km to Auckland) Auckland (10km to maunfacture site) Waimakariri River (25 km) Bromley, Christchurch (20 km) New Brighton (30 km) Canterbury (35 km) Sydney, Australia (2,600 km) North Canterbury (47 km)

study are climate change, acidification, ecotoxicity and eutrophication. These are termed midpoint categories in ReCiPe. The ability exists for ReCiPe to further classify the midpoint categories into endpoint categories such as damage to human health and ecosystem quality (Goedkoop et al, 2009). The impact assessment using ReCiPe 2008 is separated into three different perspectives; „individualist‟, „hierarchist‟ and „egalitarians‟. The „individualist‟ perspective is based on interests in the short term whereas the „egalitarian‟ is very precautionary and takes into account a long time frame. „Hierarchist‟ lies in the middle and is most commonly implemented in scientificstudies. The „hierarchist‟ perspective was selected for this study.

2.2.3 Excavated Volume The volume of soil to be excavated to construct each stormwater treatment system was analysed. Table 3 shows the difference in these excavation volumes. Excavated volume (m3) 46 3.8

Significant uncertainty exists in the endpoint results, due to the subjective nature of the analysis. The magnitude of the contribution of each category to the endpoint is not clearly defined, i.e. how much impact climate change has on human health. For this study results have been reported as midpoint categories of climate change (kg CO2 equivalents), terrestrial acidification (kg SO2 equivalents), freshwater eutrophication (kg P equivalents), marine eutrophication (kg N equivalents) and terrestrial ecotoxicity (kg 1,4 dichlorobenzene (DB) equivalents).

Addington Hynds Downstream Defender 94 TP10 Table 3. Excavated volume for treatment systems These large differences in excavated volumes affect how much each system contributes towards the impact categories. See results section for a discussion of the implications of the differences. 2.3. Impact assessment

In this context, the category „climate change‟ refers to the impact associated with the addition of carbon dioxide and other greenhouses gases such as methane and nitrous oxide to the atmosphere. All greenhouses gases emissions are converted to carbon dioxide equivalents using the model developed by the Intergovernmental Panel on Climate Change (IPCC).

ReCiPe 2008 was the method selected to conduct the impact assessment. Within SimaPro, ReCiPe gathers information about each stormwater treatment system from the ecoinvent database and groups the effects according to predefined categories. Most relevant to this

4

For example the release of 1 kg methane results in the same amount of climate change as 25 kg CO2. This gives units for climate change of kilograms of carbon dioxide equivalents. The same process is applied to the other impact categories to give their equivalents.

Assumptions All treatment systems have the same catchment area and treat the stormwater to the same level. Operation and Maintenance excluded from the model as the required information was not available. Maintenance is typically finalized after the treatment system has been implemented. Unknown transportation were assumed to be local. Limited vehicles were available with the database. Auckland Regional Council rain garden design guidelines are appropriate for Christchurch conditions. Ecoinvent database is accurate, well researched and these results are representative for New Zealand conditions The impact of no stormwater treatment is greater than operating a stormwater treatment system. ReCiPe is an appropriate impact assessment method for stormwater treatment systems in New Zealand.

The category of terrestrial acidification is concerned with the deposition of inorganic substances such as nitrous oxides, ammonia and sulphur dioxide into the soil. This results in a change in the acidity of the soil which negatively impacts biota. The category is represented by the units of sulphur dioxide equivalents. Eutrophication occurs as a result of the excessive addition of nutrients to an aquatic ecosystem, and can result in the substantial growth of primary producers. The continuation of this process over a period of time significantly depletes oxygen levels. For this model it is assumed that the limiting nutrient is nitrogen in all coastal waters and phosphorous in all freshwaters, thus they are separated into two different impact categories.

Table 3 Assumptions made whilst using Simapro metrics used. Climate change impacts were determined to be higher for all three treatment systems, than the other metrics assessed. For all impact categories shown the TP10 was the largest contributor closely followed by Hynds Downstream Defender. This result is of surprise and contradicts our hypothesis. The stormwater treatment installed in Addington was shown to have a significantly lower impact. Trends within the impact categories are discussed below.

The category of ecotoxicity refers to the impact of emissions of toxic substances to terrestrial ecosystems. Factors are normalised to kg 1,4 dichlorobenzene (DB) equivalents. 2.4. Assumptions The assumptions made throughout our project and those made in Simapro are summarized in table 3. The limitations to these assumptions are discussed below.

3.1. Climate Change 3.

RESULTS AND DISCUSSION The impact from climate change was separated into the three key processes (excavated volume, raw materials

An overview of results is located in table 4. Figure 5 shows the overall impact for each system for each of the

Table 4. ReCiPe Impact Assessment Results EV = Excavated volume, RMC = Raw Materials and Construction, TM = Transportation of Materials Climate Change (CC) (kg CO2 eq) 500

Terrestrial Acidification (TA) (kg SO2) 2.1

Freshwater Eutrophication (FE) (kg P eq) 0.00822

Marine Eutrophication (ME) (kg N eq) 0.916

Terrestrial Eco toxicity (TE) (kg 1,4 DB eq) 0.0447

EV

25

0.194

0.000196

0.109

0.00225

RMC

320

1.09

0.00577

0.378

0.0161

TM

154

0.817

0.00225

0.429

0.0263

Total

1230

5.1

0.0272

2.28

0.145

EV

2.02

0.016

1.62E-05

0.00898

0.000185

RMC

603

1.53

0.0185

0.475

0.0457

TM

622

3.55

0.00873

1.8

0.0989

11700

94.7

0.0618

12.7

0.229

50

0.397

0.000401

0.223

0.0046

11,100

91.4

0.0534

11

0.13

548

2.91

0.00804

1.52

0.094

Hynds Downstream Defender

Addington

Total

TP10 Rain Garden Design

Total EV RMC TM

5

and construction, transport) for each of the treatment systems (shown in figure 6). The process with the largest contribution to climate change was shown to be raw material and construction. The difference of impact between the TP10 and Hynds Downstream Defender, from the process of raw materials and construction, is significant in that the contribution of the Hynds is 603 kg C02 eq whilst the TP10 contribution is 11,100 kg C02 eq. The contribution towards climate change from the transportation process for the Hynds downstream defender and TP10 is very similar, with the Hynds contributing 622 kg C02 eq whilst the TP10 contributes 548 kg C02. This similarity in transport contribution was surprising since the components of the systems Hynds Downstream Defender are sourced from a large distance (as shown in table 2). Thus it was important to conduct further analysis

ReCiPe Impact Categories 100000 Addington

10000

Hynds 1000

kg - eq

TP10 100 10 1 0.1 0.01 0.001 CC - (CO2)

TA - (SO2)

FE - (P)

ME - (N)

TE - (1,4 DB)

Figure 5. ReCiPe impact categories CC = Climate change, TA= Terrestrial acidification, FE = Freshwater eutrophication, ME = Marine eutrophication, TE = Terrestrial ecotoxicity (note: logarithmic scale)

Through investigation it was determined that, although the overall distance travelled is important, in determining the climate change impact from transportation it is vital to assess the proportion of the distance travelled by both road and ocean. The results of this assessment are shown on figure 7. In this figure the left vertical axis represents the transportation contribution for each system. Road and ocean transport are separated to represent the proportion that contributes to the overall impact. The right hand axis of figure 7 shows the climate change impact associated with transportation for all three treatment systems. Analysis shows that there is a noticeable difference in the ocean transportation amount between the Hynds Downstream Defender (2,500 tkm) and the TP10 rain garden (1,100 tkm). This difference is not reflected in the climate change impact results (Where the TP10 is 548 kg CO2 eq compared to Hynds which is 622 kg CO 2 eq). This shows that the impact upon climate change is primarily due to the proportion of distance travelled on the road.

Figure 6. Impact on climate change for three treatment systems (note: logarithmic scale)

Transportation Climate Change Impact

6000

Ocean

600

Tonnes * kilometers (tkm)

5000

Road

500

Climate Change

400

4000 3000

300

2000

200

1000

100

0

0

Addington

The process shown as „excavated volume‟ represents the removal of the initial substrate at the site to allow for the construction of the stormwater treatment. This was modelled as being carried out by a hydraulic digger. The key factor that influences the impact of this process is the size of the stormwater treatment system. The

Hynds

TP10

Figure 7. Proportion of road and ocean transport for three stormwater systems against climate change impact.

6

700

Climate Change impact (kg CO2 eqv)

The raw materials used in the TP10 design were shown to have a higher climate change impact than all other systems (figure 6). Through further analysis it was determined that 98% of this impact was produced by the 33 tonnes of compost utilized in the design. The impact was a lot higher than expected, so the information used in the calculation was assessed. The information within the ecoinvent database used in the model originates from Switzerland in 1997 (Nemecek and Kagi, 2007). Conditions and technology available in Switzerland at this time were very different to New Zealand conditions currently. It is expected that if the database was updated the resulting impact upon climate change would be significantly reduced.

Hynds downstream defender is significantly smaller than the TP10 system, which explains the significant difference in the impact from „excavated volume‟ see table 3. 3.2. Terrestrial Acidification Information about the contribution of each life cycle process (Excavated volume, raw materials and construction, and transportation) to terrestrial acidification is shown on figure 8. Trends for this impact category are similar to those shown for climate change, although it is important to note that the impact is much smaller in magnitude. Hynds transportation has a contribution of 3.5 kg SO2 eq compared to the climate change impact of 622 kg CO2 equiv. Raw materials for the TP10 design are significantly higher (91.4 kg SO2) than the Hynds system (1.53 kg SO2) and Addington system (1.09 kg SO2). This is attributed to the technology involved in producing the compost for this design.

Figure 8. Impact on terrestrial acidification for three treatment systems (note: logarithmic scale)

3.3. Freshwater Eutrophication The process contributions of the three stormwater treatment systems to freshwater eutrophication are shown graphically in figure 9. The impacts for this category are much lower than all other categories assessed (see figure 5). Raw materials and construction are shown to have the largest contribution, a surprising result. The impact from transportation is attributed to phosphorus being adsorbed to sediment which accumulated on the roadway. This sediment is then washed off the road by rainfall. It is interesting that this impact is produced by the construction of a stormwater treatment system as the system is designed to remove suspended sediment.

Figure 9. Impact on freshwater eutrophication for three treatment systems (note: logarithmic scale)

3.4. Marine Eutrophication Impacts from the three treatment systems on marine eutrophication are shown in figure 10. The impacts are significantly higher than the impacts from freshwater eutrophication. This is clearly shown by the Hynds transportation impact which varies from 1.8 kg N eq to 0.00873 kg P equiv. The contribution from raw materials and construction of the TP10 design are shown to be significantly higher (11 kg N eq) than both the Hynds (0.475 kg N eq) and Addington systems (0.378 kg N eq). This is also attributed to the compost database information, as nitrogen is known to be a component of the compost (Nemecek and Kagi, 2007). Figure 10. Impact on marine eutrophication for three treatment systems (note: logarithmic scale)

3.5. Terrestrial Ecotoxicity The process contributions of the three stormwater treatment systems are represented graphically in figure 11. The contribution from the transportation process for the Hynds Downstream Defender (0.0989 kg 1,4 DB eq)

7

ReCiPe Impact Categories 100000 Addington

10000

Hynds TP10

1000

kg - eq

TP10 (60% size) 100 10 1 0.1 0.01

Figure 11. Impact on Terrestrial ecotoxicity for three treatment systems (note: logarithmic scale)

0.001 CC - (CO2)

A sensitivity analysis was conducted in order to investigate the sensitivity of the results to the input parameters. Two factors were addressed; the size of the TP10 rain garden and the transportation associated with the Hynds Downstream Defender.

In order to simulate the impact from decreasing the distance from the source of the materials to the site the transportation impacts for the Hynds Downstream Defender were decreased by 50%.

It has been shown that the rain gardens sized using the TP10 are typically overdesigned (Smythe et al, 2007). In order to investigate the impact this has on the categories addressed input values were determined for a rain garden 60% smaller than the original design. The resulting surface area is 33 m2. The result this change has on the impact categories assessed is shown on figure 12.

Climate Change (kg CO2 eq) Terrestrial Acidification (kg SO2) Freshwater Eutrophication (kg P eq) Marine Eutrophication (kg N eq) Terrestrial Ecotoxicity (kg 1,4 DB eq)

% Change

11700

7030

-39.9

94.7

57.0

-39.8

TE - (1,4 DB)

4.2. Hynds Transportation

4.1. Rain Garden Size

TP (40% decrease)

ME - (N)

kg 1,4 DB eq) just below that of the Hynds Downstream Defender (0.145 kg 1,4 DB eq).

SENSITIVITY ANALYSIS

TP 10 Original Size

FE - (P)

Figure 12. Impact of reduced TP10 rain garden size (60% of original) (note: logarithmic scale)

is shown to be similar to the contribution of the TP10 rain garden (0.094 kg 1,4 DB eq). 4.

TA - (SO2)

Hynds original Climate Change (kg CO2 eq) Terrestrial Acidification (kg SO2) Freshwater Eutrophication (kg P eq) Marine Eutrophication (kg N eq) Terrestrial Ecotoxicity (kg 1,4 DB eq)

% Change

622

Hynds (50% decrease) 311

3.55

1.78

-49.9

0.00873

0.00437

-49.9

1.8

0.9

-50.0

0.0989

0.0494

-50.1

-50.0

Table 5, Sensitivity of 50% decrease in Hynds Transportation 0.0618

0.0371

-40.0

12.7

7.66

-39.7

0.229

0.137

-40.2

The result of the reduction of transport distance was a 50% decrease in impact across all categories (table 6). In theory the materials required for the Hynds downstream defender can be sourced closer to the Addington site, this reduction in transportation has a significant effect on the environmental impact of the system.

Table 4, Sensitivity of TP10 size (40% decrease) The reduction in size is shown to have a significant impact across all metrices analysed as approximately a 40% percent reduction in impact was observed (table 5). This shows the impact correlates directly to the size of the system installed. In the case of terrestrial ecotoxicity the reduction in size of the TP10 brings the value (0.137

5.

RECOMMENDATIONS

Information used in the study was based on European conditions. Utilising data more specific to a New Zealand context would improve the accuracy of the results. The Life Cycle Association of New Zealand (LCANZ) is currently developing a database specific to

8

New Zealand conditions. It is recommended that the LCA is repeated using this database to attain results specific to New Zealand.

implementation methods for stormwater treatment systems, to mitigate environmental impacts. The possibility exists to mitigate climate change impacts through the addition of plants to the treatment system (in the case of the vegetated rain garden) or by utilizing clean burning fuels in transportation. It is important to note that greater restrictions are placed on vehicle emissions in Europe than New Zealand, thus the impacts of New Zealand vehicles are likely to be greater than those utilized in the model. Impacts upon freshwater eutrophication and terrestrial ecotoxicity from the construction of a stormwater treatment system are complex to mitigate. This must be an important consideration when selecting an appropriate “environmentally friendly” stormwater treatment system.

Once a more robust model is developed using New Zealand specific data, it is recommended that a greater range of stormwater treatment systems are modelled. Conducting life cycle assessments for a range of treatment options will quantify the environmental impacts. This will provide a valuable tool to inform stormwater management decisions. Modelling the impact of no stormwater treatment would provide an interesting comparison to the impacts of installing a stormwater treatment system. This is not possible with the current LCA framework considered in this study. The impacts of no stormwater treatment system are assumed to be much greater than the environmental impact of installing a treatment system. Further developments to the model would aid to quantify this claim. A significant factor is the output water quality, and the long term impacts this has on the natural environment. 6.

7.

ASSOCIATED CONTENT

A table of supporting information associated with the model development is available free of charge via the internet at http://www.hydroeco.info/stormwaterLCA 8.

CONCLUSION

ACKNOWLEDGEMENTS

Many thanks to Tyler Hengen at South Dakota School of Mines and Technology for his help with SimaPro 7.3 also Thierry Amat and Mike Strickett at Hynds Environment for information about the Downstream Defender. SimaPro 7.3 software was provided by the Civil and Natural Resources Engineering Department.

The environmental impact associated with three different stormwater treatment systems were compared using a life cycle assessment. The three systems compared were a subsurface sand filter currently installed in Addington, Christchurch, a vegetated rain garden design using NZ guidelines and a proprietary Hynds Downstream Defender. Impact categories of climate change, terrestrial acidification, freshwater eutrophication, marine eutrophication and terrestrial ecotoxicity were addressed. The TP10 rain garden was determined to have the greatest impact across all categories assessed. This was closely followed by the rain garden designed with the Hynds Downstream Defender. The Addington system was determined to have the lowest impact. Climate change impacts were significantly greater than all other impact categories.

9.

REFERENCES

Althaus, H.-J., Doka, G., Dones, R., Heck, T., Hellweg, S., Hischier, R., et al. (2007). Overview and methodology of ecoinvent data v 2.0. Switzerland: Swiss centre for life cycle inventories. Amat, T. (2011). Personal Communication. Hynds Environmental. Andrew, R., & Vesely, E.-T. (2008). Life-cycle energy and CO2 Analysis of stormwater treatment devices. Water Science and Technology, 58(5), 985-993. ARC. (2003). Stormwater management devices: design guideline manual. Techincal Publication 10, Chapter 7: Auckland Regional Council. CCC. (2003). Waterways, wetlands and drainage guide. Christchurch, New Zealand: Christchurch City Council. Frischknecht, R., Jungbluth, N., Althaus, H.-J., Doka, G., Dones, R., Heck, T., et al. (2007). Ecoinvent report no 1, Chapter 1 Overview and Methodology. Switerland: Swiss centre for life cycle inventories. Goedkoop, M., Heijungs, R., Huijbregts, M., & Schryver, A. (2009). ReCiPe 2008. Netherlands: Ministry of Housing, Spatial Planning and Environment. Goedkoop, M., Oele, M., Marisa, A., & Hegger, S. (October 2010). Simapro database manual

The selection of raw materials and transportation distances were shown to influence the magnitude of impact categories. It was shown that road transportation had a large contribution to the magnitude of the climate change impact, especially for the Hynds Downstream Defender. The compost material used in the TP10 rain garden was shown to have a significant contribution to all impact categories. This is attributed to the 1997 data for this material, which originates from Switzerland. It is deemed this data is out of date and not representative of New Zealand conditions and technology. Variations were made to both the raw materials used in the TP10 rain garden and the transportation of the Hynds Downstream Defender to assess the influence upon impact categories. For both cases it was shown to decrease all impact categories. This provides good insight into how best to modify designs and

9

methods maunal. Netherlands: PRe Consultants. Good, J. (2011). Water quality treatment and hydraulic efficacy of laboratory and field rain garden. Unpublished master's thesis for master's degree, University of Canterbury, Christchurch, New Zealand. Griffiths, G., Pearson, C., & McKerchar, A. (2009). Review of the frequency of high intensity rainfalls in Christchurch. Christchurch: NIWA. Hynds Environmental. (2006). Hynds Downstream Defender. Retrieved June 9, 2011, from Hynds Environmental: http://www.hyndsenv.co.nz/ViewProduct.aspx ?Id=12 ISO. (2006). 14040: Environment Management - Life Cycle Assessment Principles and Framework. International Standardization. Kirk, B., Roseen, R., & Etnier, C. (2006). The Big Picture - Evaluating Stormwater BMPs Through the Life Cycle Lens. University of Vermont. Nemecek, T., & Kagi, T. (2007). Life Cycle Invertories of Agricultural Production Systems. Switzerland: Ecoinvent centre. PRe Consultants. (2010). Introduction to LCA with Simapro 7. Netherlands: Pre Consultants. Sansalone, J., & Buchberger, S. (1997). Characterization of solid and metal element distributions in urban highway stormwater. Water Science and Technology 36(8-9): 155160. Smythe, C., Diyagama, T., Carter, B., & Servaas, B. (2007). A raingarden with a difference. Auckland: North shore City Council. Spielmann, M., Bauer, C., Dones, R., Villigen, P., & Tuchschmid, M. (2007). Transposrtation services ecoinvent data v2.0. Switizeland: Swiss centre for life cycle inventories. Wimmer , W., Zust, R., & Lee, K.-M. (2004). Ecodesign Implementation: A Systematic Guidance on Integrated Environmental Considerations into Product Development. Cambridge: Springer.

10