Eco-Efficiency Assessment Utilizing Recycled Glass Aggregate in Concrete

Abstract

This paper reviews specific technical and eco-efficiency performance issues in using glass waste as an aggregate in the production of concrete. Eco-efficiency is a relatively modern tool in the pursuit of sustainability. Eco-efficiency is the concept of maximising the benefits from the use of non-renewable resources while minimising the use of non-renewable resources. The paper details a life cycle assessment and eco-efficiency review of a potentially sustainable alternative to traditional concrete, made from ordinary Portland cement. The study follows the ISO framework, which includes goal and scope, a life cycle inventory, life cycle impact assessment, life cycle costing, normalising of data and the creation of an eco-efficiency portfolio. SimaPro life cycle assessment software has been used to further analyse the use of recycled glass aggregate as a replacement for naturally occurring stone aggregate in geopolymer concrete. The study found that the use of geopolymer concrete as a non-cement based alternative concrete was a viable way to reduce emissions with a high global warming potential but faced challenges in other environmental impact areas. There is a need for ongoing research and study on the application of eco-efficiency as a tool in the pursuit of sustainable practices in society.

1. Introduction

The construction industry in Australia is the third largest industry in terms of the total number of people employed as well as the share of GDP in Australia, employing over 1.15 million people and generating 360 billion dollars in revenue in 2019 [1]. Construction work is resource and energy intensive; it involves heavy machinery and trucks transporting large volumes of material contributing 18.2% of the total Australian carbon footprint [2]. This makes it a suitable area of focus to improve sustainability and reduce carbon emissions. Studies have indicated that upwards of 82% of the total carbon emissions from the construction stage of buildings is due to construction materials [3]. Concrete is a key construction material that typically uses large volumes of virgin sand and aggregate mixed with water and ordinary Portland cement, an energy intensive material to produce. This makes the production and use of concrete in Australia a suitable area of interest to apply eco-efficient solutions to minimise material use and emissions.

An estimated 0.9 ton of greenhouse gas emissions are released to produce one ton of ordinary Portland cement [4], making cement the dominant contributor to environmental degradation in the production of concrete. Cement production is estimated to contribute 5–8% of all man-made carbon dioxide emissions [5]. By incorporating alternate ingredients when producing concrete, the total lifetime emissions can be reduced, producing a more eco-efficient concrete. Any alternative to traditional Portland cement concrete must have a comparable price, emissions, consistency and strength class. Using industrial waste, such as fly ash (FA) and silica fume (SF), as binding agents in place of cement provides an added benefit of reduced landfill costs otherwise associated with those materials, apart from using construction waste or other recycled aggregates as a replacement to virgin natural aggregates.

The use of virgin materials has been increasing steadily throughout history in tandem with the economic growth of society. Although the global extraction rate of natural resources has increased eight times as much on average from 1900 to 2005, the extraction of construction materials has grown at a much higher rate: thirty-four times higher [6]. By the year 2050, it is estimated that humans could be consuming 140 billion tons of non-renewable resources per year in the form of fossil fuels, minerals and ores [6]. This rate of virgin resource use is not sustainable and must be uncoupled from the economic growth of society to ensure that it can continue without causing irreparable damage to the natural environment. Improving the level of resource productivity throughout society can be achieved by maximising economic output while minimising the environmental impacts. Recycling is a major strategy to minimize waste streams [7]; converting waste into useful material streams, such as geo-polymer concretes, can reduce the demand for virgin resources as well as reduce disposal costs associated with landfill.

Recycled glass has the potential to be used as an aggregate replacement and a constituent of concrete. The current rate of recycling glass packaging is estimated to be approximately 46% in Australia, so there is a need to find applications for recycled glass packing to increase the diversion rate from the landfill [8].

Glass is a material highly suitable for recycling as it is commonly produced and used, chemically inert and waste collection processes already exist. Crushed glass is a product of the recycling industry, obtained after crushing glass from a variety of end-use sources. The chemical composition of glass changes depending on the initial application, with most of the glass packaging produced being soda–lime–silicate glass [9]. Although glass can be melted and reformed into new packaging, small amounts of contamination including ceramics, metals and mixed glass colours make the glass unsuitable for remelting as glass packaging. The ability of the construction industry to tolerate higher levels of contamination and take large volumes of crushed glass makes it a suitable solution for diverting glass waste from landfill [8].

It is important to use scientific tools to assess the sustainability of using waste glass as aggregate in concrete, in addition to the technical feasibility. Life cycle assessment (LCA) tools incorporate all aspects of a product from cradle to grave, including as many financial and environmental impacts as possible to represent the true impact over the life of a product or process. Eco-efficiency reviews the environmental costs so that they can be weighed and compared against the financial benefits to prevent the environmental limitations of the Earth being exceeded. There is a lack of current literature using life cycle assessment techniques and eco-efficiency reviews to assess the viability of using recycled glass as an aggregate replacement in concrete. The benefits of eco-efficiency studies demonstrate that there is a need for eco-efficiency reviews on the use of recycled glass aggregate to produce concrete in the construction industry. An eco-efficiency review is a tool used to find a suitable compromise between the financial demands of businesses and individuals and the environmental limitations of the Earth, to ensure that proposed solutions such as using recycled glass aggregate in construction concrete are sustainable [10].

A study on the development of eco-efficient bricks used an LCA approach to determine the eco-efficiency of several different designs for eco-efficient interlocking bricks and compare them to conventional bricks [11]. It found that most of the interlocking brick designs reviewed could be considered eco-efficient. Another Curtin University study on eco-efficiency involved a sustainability assessment of recycled aggregate concrete mixes containing industrial by-products [12]. Experimental results have shown that the new concrete formulations may be suitable for higher strength compressive applications alongside residential flooring and other lower strength applications. The system boundaries included all stages from cradle to grave with environmental impacts similar to those identified by Biswas and Zhang [13]. The study is comprehensive and is a model example of how an eco-efficiency analysis should be conducted. It easily contrasts each alternate option to determine the most effective solution. The study was unable to improve both economic and environmental aspects, hence potential improvements would include reducing this limitation.

Continuing research into green alternatives in the construction industry of Australia is essential for advancing incremental change towards more sustainable solutions. The civil construction industry generates 40–50% of all the global output of greenhouse gas emissions [14], with construction and demolition waste representing a portion of this total impact. LCA studies in Australia can work to analyse the entire life cycle, i.e., from extraction of components, production of concrete to end of life disposal and demolition. Changing technologies, resource extraction methods, electrical generation mixtures and end-of-life processing can impact the areas of concern when studying the LCA impacts of concrete. Changing costs due to societal pressures and expectations can also influence the most cost effective, environmentally beneficial solution. Current studies in Australia suggest that the use of recycled aggregate in concrete as a replacement for virgin aggregates has the potential to be a viable solution for reducing environmental impacts [15]; however, care must be taken to fully understand the true impact of transporting and crushing recyclable materials. Future LCA studies should continue to develop with increasing complexity, including all stages of the life cycle (cradle-to-grave), to increase the accuracy of the models, promoting change where possible. No research to date has conducted the eco-efficiency analysis of concrete made from recycled glass. This research is important because the cost of recycling could outweigh the environmental benefits of the use of recycled glass in concrete or vice versa. This is still unknown. As alluded earlier, eco-efficiency has been used to determine cost-effective and environmentally friendly options for the construction sector. This article partially uses the information of the companion paper of this issue to assess the eco-efficiency performance of the recycled glass-based concrete [16]

2. Materials and Methods

The purpose of the study is to assess the viability of using recycled glass aggregate (RGA) as a virgin aggregate replacement in the production of geopolymer concrete. To assess the benefits of using RGA, geopolymer concrete samples have been produced using increasing amounts of RGA. The samples will be compared to a geopolymer concrete sample that contains no RGA with the same amount of binder being used. A life cycle assessment (LCA) will be produced in accordance with the ISO 14040 guidelines [17]. The functional unit used in the life cycle inventory (LCI) to compare each of the samples will be 1 m³ of concrete and the transportation measurement will be tonne-kilometres, with separate databases for road and sea transport. The life cycle environmental impact assessment (LCIA) will consider the cost per strength (i.e., AUD/MPa) as the functional unit.

The LCA consists of four stages: goal and scope, a life cycle inventory (LCI), a life cycle environmental impact assessment (LCIA) and an interpretation of the LCA. The goal of the LCA is to compare the degree of eco-efficiency of each of the concrete samples produced with varying amounts of RGA. The scope involves sourcing of raw material, manufacturing, transport and construction at Curtin University (cradle-to-gate). The scope is specific to the samples produced at Curtin University and does not include potential real-world application of the concrete. The LCI will be produced to estimate the inputs of each of the samples, using data specific to the samples as produced at Curtin University with Australian conditions. The normalisation factor is based on a normalisation list for Australian specific values [18]. The data will be normalised and weighted against Australian specific values to create an eco-efficiency portfolio which will be discussed in the interpretation.

2.1. Goal and Scope

The methodology for the goal and scope involves setting system boundaries and determining definitions for elements of the system. Changing system boundaries will change the overall impacts due to the items to be assessed: larger system boundaries will create more complex, inclusive systems and smaller system boundaries will simplify systems, making calculations easier and more accurate. Although there are benefits to larger systems, the increased modelling complexity can make a comparison between two systems more difficult. An example can be seen when comparing two different manufacturing techniques for an identical item. The system boundaries can be set to exclude the life and disposal of the product, as the results can be assumed to be identical for identical products.

Critical units must be determined during the goal and scope phase to be able to compare and contrast different products and processes. Almost all standard units that exist can be used, if they are consistent across the LCA comparison. Typically, units of volume, strength or mass would be suitable for use in defining the LCA parameters,

The structure of the LCA will follow internationally recognised standards, specifically the International Organization for Standardization (ISO), ISO (2006) 14.040 Environmental management life cycle assessment principles and frameworks [17]. This standard provides guidelines for a life cycle assessment, specifically in defining the four stages, including goal and scope, inventory analysis, impact assessment and interpretation, as seen in Figure 1.

2.2. Life Cycle Inventory (LCI)

The LCI involves determining the constituents involved in the creation of the product, service or process to be assessed during the LCA. This is a necessary step to determine the environmental impacts during each stage of the production of raw materials, transport, energy consumption, manufacturing, use and final disposal (including recycling). Using the critical units determined during the goal and scope (e.g., 1 kg of steel plates), an LCI can be created that contains all the estimated inputs involved in the creation of that specific item being assessed. The importance of this step is to define the exact inputs going into the creation of the item at a specific time and place. The processes and transport methods for identical items created in different countries can change the overall impacts produced as a result, so it is important to specify exact material acquisition methods, transport methods, distances and processes.

2.3. Life Cycle Impact Assessment (LCIA)

The purpose of the LICA is to gain an understanding of the total environmental impacts resulting from the production of one functional unit of the focus material. The environmental impacts chosen are the direct consequences of the release of certain chemicals into the environment or their equivalent. Global warming potential is a serious threat to natural ecosystems on the planet. Releases of CO₂ during the life cycle of concrete can be measured, while the release of other chemicals such as methane, a more potent greenhouse gas, has an equivalency to CO₂ that can be calculated. SimaPro software uses the ReCiPe 2016 method to determine life cycle impacts, drawing information from known databases of information.

A collection of environmental impacts, seen in

Table 1. Impact assessment methods and normalisation factors [18].

Environmental Impacts	Gross Domestic Environmental Impact	Weighting
Global warming potential	28,690 kg CO2 eq	19.50%
Eutrophication	19 kg PO43− eq	2.90%
Water depletion	930 m3 H2O	6.20%
Land use and ecological diversity	26 Ha a	20.90%
Photochemical smog	75 kg NMVOC	2.80%
Human toxicity	3216 kg 1,4-DB eq	2.70%
Terrestrial ecotoxicity	88 kg 1,4-DB eq	10.30%
Freshwater ecotoxicity	172 kg 1,4-DB eq	6.90%
Marine ecotoxicity	12,117,106 kg 1,4-DB eq	7.70%
Ionising radiation	1306 kg U235 eq	1.90%
Ozone depletion	0.002 kg CFC-11 eq	3.90%
Abiotic depletion	300 kg Sb eq	8.20%
Acidification	123 kg SO2 eq	3.10%
Respiratory inorganics	45 kg PM2.5 eq	3.00%

, lists the environmental impacts specific to Australian conditions. Each of these environmental impacts is associated with a specific causal factor (e.g., CO₂ causes global warming potential). Normalisation and weighting of data are critical steps in correctly analysing the results produced during the LCA. Normalisation of data for Australian conditions involves determining the gross domestic environmental impacts for the average Australian person. The environmental impacts can then be measured in terms of equivalency to the environmental impacts of an Australian inhabitant per year. The process of normalisation is essential to understanding the relative significance specific to a geographic area or population and provides a localised impact in the correct context. The weightings associated with each environmental impacts are a measure of the relative importance of each of the individual impact. The weightings are specific to Australian conditions [18], as different impacts will have different potential outcomes on humanity.

2.4. Life Cycle Cost (LCC)

A LCC) is the concept of determining the total financial costs involved during all stages of the life cycle assessment, as determined by the goal and scope of the LCA. The true costs determined through the production of an LCC may involve costs from planning, producing, operating, maintaining and disposing of an asset, product or process. The costs will be determined from the collection of constituent components, identified in the LCI stage of the LCA. It is critical that all associated costs for each individual component are accounted for during the LCC stage, including raw material acquisition, transport, production, usage and disposal. The collection of data specific to the product being assessed can increase the accuracy of the modelling at the expense of far more complex models that can be harder to manage. Simplification of costs may involve standardised data for elements such as transport, using a predetermined cost per km approach as opposed to the specific costs associated with the production of an individual product. It is important to specify any assumptions made and state the methods used during the LCC stage.

2.5. Normalising Data Values to Produce an Eco-Efficiency Portfolio

To produce an eco-efficiency portfolio, the normalised environmental impacts must be compared against normalised economic data for the LCA [19]. The initial economic measurements from the production determined in the LCI and LCC can also be used alongside other known relevant data, such as strength. To normalise economic data, relevant GDP per inhabitant must be calculated, using accurate data of a country’s GDP and population.

To normalise the economic impacts associated with the LCA, the LCC of each sample is determined through the costing estimates of each of the constituent components. As each sample has a similar binding composition, the only difference is the amounts of natural aggregate and glass aggregate, varying the costings. The normalised economic cost (Nc) is calculated from the concrete sample cost (CS), divided by the GDP per inhabitant (Equation (1)).

$$N_c=\frac{CS}{GDP} .$$

(1)

To normalise the environmental impacts of each of the concrete samples, the fourteen individual impacts must be converted into terms of the equivalent impact of an Australian inhabitant. The values for each of the environmental impacts (EI) are divided by the gross domestic environmental impact (GI) for each specific unit of measurement, to give a normalised value, Ne (Equation (2)).

$$N_e=\frac{EI}{GI} .$$

(2)

This normalised value for each of the individual environmental impacts can be refined further with individual significance weightings, as seen in Table 1. The impact weightings (Wi) are specific to Australian conditions and determine the relevant importance of the environmental impacts of each individual during the period of a year. The sum of all weighted normalised values is the normalised environmental impact (Ee) (Equation (3)).

$$E_e={\displaystyle \sum }{N}_e\times W_i.$$

(3)

The initial positions (iPP) for the eco-efficiency portfolio are determined by the ratio of the normalised cost and environmental impact for each of the samples, as compared to the average normalised cost and environmental impact, where MX represents each of the samples M0–M4, and the average normalised environmental impacts and costs are the averages for all the samples M0–M4 (Equations (4) and (5)).

$$iP{P}_{MX, e}=\frac{E_e{}_{MX}}{E_e{}_{AV}},$$

(4)

$$iP{P}_{MX, c}=\frac{N_c{}_{MX}}{N_c{}_{AV}}.$$

(5)

The portfolio positions can be adjusted further by determining the cost-effectiveness of each sample, relative to the other samples. This changes the portfolio position of each of the samples, relative to the other samples, as the cost or environmental impact of any individual sample changes. As the performance of one type of concrete changes, the other options may become better or worse, relative to the scale of the changes. This is known as the environmental to cost relevance factor R (Equation (6)).

$$R=\frac{E_e{}_{AV}}{N_c{}_{AV}}.$$

(6)

To calculate the final portfolio (PP) positions, the normalised costs and environmental impacts must be balanced (Equations (7) and (8)) [19]. This is an important step to ensure that each of the portfolio positions to be compared are being compared relative to each other. As the parameters on one sample change (e.g., through increased emissions or costs), other samples can become better options, relative to that sample.

$$P{P}_{MX, e}=\frac{iP{P}_{AV, e}+\left(iP{P}_{MX, e}-iP{P}_{AV, e}\right)\ast \sqrt{R}}{iP{P}_{AV, e}},$$

(7)

$$P{P}_{MX, c}=\frac{iP{P}_{AV, c}+\left(iP{P}_{MX, c}-iP{P}_{AV, c}\right)/\sqrt{R}}{iP{P}_{AV, c}}.$$

(8)

2.6. Eco-Efficiency Portfolio

An eco-efficiency portfolio is a visual representation of the comparison of the level of eco-efficiency of different options available to a company. The diagram compares the normalised economic costs and normalised environmental impacts, as described in Equations (7) and (8) to determine the portfolio positions (PP) of each of the options to be compared. Options with a low eco-efficiency have higher environmental impacts relative to the costs; they are positioned on the left side of the diagonal line of the eco-efficiency portfolio. Any option to the right side of the diagonal line on the eco-efficiency portfolio is considered to have higher eco-efficiency, representing a ratio of lower environmental impacts relative to the economic costs. An example of an eco-efficiency portfolio can be seen in Figure 2. The diagonal line represents the ratio of PPMX, e and PPMX, c, where any point on the line is equal to one. Points in the eco-efficiency portfolio where the ratio is ≥1 represent an eco-efficient option (Figure 2).

3. Results

3.1. Goal and Scope

The goal of the LCA study is to compare environmental impacts of alternate concrete samples produced at Curtin University. There are five concrete samples, each containing a different ratio of recycled glass aggregate to virgin natural aggregate. The concrete samples do not use traditional Portland cement as binder, but instead use an alternate binder composed of fly ash, sodium silicate and sodium hydroxide. The LCA will be used to produce normalised data for Australian specific economic and environmental conditions to produce an eco-efficiency portfolio that compares the eco-efficiency each of the samples. The recycled glass aggregate concrete mixtures have the potential to be used in structural applications, requiring a compressive strength of 40 MPa. The benefit of the eco-efficiency review is the inclusion of economic factors which can be a major influence in the decision to change construction processes to be more sustainable. The application area of the LCA and subsequent eco-efficiency review is the construction industry in Australia, where there is potential for alternatives to traditional concrete made with Portland cement and virgin aggregate, namely concrete made with recycled glass aggregate.

Several critical units of measurement have been determined. The unit of transport chosen for emission and economic estimates is the tonne-kilometre (t-km). This unit takes into consideration the impacts that weight has on transport and processing of materials, alongside the distance of transport. Reliable databases exist for t-km measurements, as it is a typical unit of measurement in LCA software, and in SimaPro specifically. The functional unit of measurement of each mixture has been decided at 1 m³. This volume of concrete will be used to determine the environmental and economic factors for each of the samples and can be easily scaled up to estimate larger volumes of concrete. Although the samples may have variations in density, the essential characteristic of concrete is its load bearing compressive strength. This characteristic makes a volumetric measurement of concrete a suitable quantifiable unit of comparison. To compare the economic differences between the concretes, strength tests will be conducted on samples made from each mixture. These strength values compared to the cost of each mixture will determine the cost per MPa of strength (AUD/MPa).

The products chosen are the five concrete samples produced at Curtin University, (M0–M4). The system included in the analysis is a “cradle-to-gate” approach, including the mining or collection of raw materials, processing, transport, mixing and production of the samples. The study will not include the “gate-to-grave” portion of the life of the samples. This decision has been made to limit the complexity of the initial modelling in this study and has been deemed a suitable approach. The assumption is that if used at scale in industry, the alternate concrete mixtures, as represented by the five Curtin University samples, would have very similar life cycle impacts from gate to grave as compared to traditional concrete mixtures. The same methods would be used to transport the concrete, erect structures, and the subsequent end of life construction and demolition waste procedures. Future studies on the different end of life disposal procedures have potential to be conducted.

3.2. Life Cycle Inventory (LCI)

The LCI is a determination of the materials, processes and transport that contribute to the creation of a product. The concrete samples at Curtin University, labelled M0–M4, each contain materials as described in

Table 2. Mix designs of one cubic meter of concrete in kg/m³.

		Fly Ash	Sand	20 mm NA	10 mm NA
Mix no	%GA	kg/m3	kg/m3	kg/m3	kg/m3
M0	0% GA	400	684	592	592
M1	10% GA	400	684	532.8	532.8
M2	20% GA	400	684	473.6	473.6
M3	30% GA	400	684	414.4	414.4
M4	40% GA	400	684	355.2	355.2
Mix no	13–19 mm GA	7–13 mm GA	Sodium silicate	* Sodium hydroxide	Water *
M0	kg/m3	kg/m3	kg/m3	kg/m3	kg/m3
M1	0	0	114.3	10.83	34.87
M2	59.2	59.2	114.3	10.83	34.87
M3	118.4	118.4	114.3	10.83	34.87
M4	177.6	177.6	114.3	10.83	34.87
	236.8	236.8	114.3	10.83	34.87

* Undiluted SH and water combine to 8M sodium hydroxide solution.

. The sand and natural aggregate is typical of concrete samples. The fly ash is specific to concrete samples using non-cement binder and requires the use of sodium silicate and sodium hydroxide to react with the fly ash, thus curing the concrete. The amount of glass aggregate (GA) in each sample changes, as the purpose of the samples is to compare the compressive strengths of each sample with a different ratio of recycled glass aggregate. As the functional unit is 1 m³, the raw material requirements contributing to 1 m³ were determined for each mixture These values were provided by Curtin University from the production of each concrete sample and are assumed to be accurate and correct.

The LCI of each of the concrete mixtures has been further developed, as seen in

Table 3. Life cycle inventory of concrete mixtures.

Mix Number	M0 (0% Glass)	M1 (10% Glass)	M2 (20% Glass)	M3 (30% Glass)	M4 (40% Glass)
Mix ID	0PC	10RGA	20RGA	30RGA	40RGA
Fly ash (kg)	400	400	400	400	400
Sand (kg)	684	684	684	684	684
NA 20 mm (kg)	592	532.8	473.6	414.4	355.2
NA 10 mm (kg)	592	532.8	473.6	414.4	355.2
GA 13–19 mm (kg)	0	59.2	118.4	177.6	236.8
GA 7–13 mm (kg)	0	59.2	118.4	177.6	236.8
Alkali solution (kg)	160	160	160	160	160
Total weight (kg)	2428	2428	2428	2428	2428
Road transportation (tkm)	56.72	60.51	64.30	68.09	71.88
Sea transportation (tkm)	1142.4	1142.4	1142.4	1142.4	1142.4
Manufacturing (kWh)	81.7	81.7	81.7	81.7	81.7
Transportation distance km		Road		Sea
Gladstone Eraring		19		2856
Hanson, Lexia		32
Holsim Gosnells		18
Holsim Gosnells		18
Spartel PTY LTD		50
Spartel PTY LTD		50
Coogee Chemicals		37

. The origin of each of the constituent components is listed on the right side of the table, allowing values for transport for sea and road to be calculated. The LCI is specific to the samples produced at Curtin University, with the destination being Curtin University where the samples were produced. In future construction involving recycled glass aggregate concrete, the transport distances would change as construction locations change. Specific LCA studies could be conducted for individual construction projects; however, the results from this study will be beneficial in determining the suitability of using concrete produced with recycled glass aggregate.

3.3. Life Cycle Impact Assessment (LCIA)

The environmental impacts of each of the concrete samples (M0–M4) were determined using the life cycle inventory produced in Table 3. Each of the constituent materials used to produce the samples have their own environmental impact from the mining, transport and construction of the concrete samples. The functional unit of 1 m³ of concrete was used for each sample, as described in the LCI. Existing databases in SimaPro software were used at Curtin University to calculate the fourteen environmental impacts considered relevant to Australian conditions, as seen in Table 1. The databases used included the Australian indicator set, the Traci method, the ReCiPe method and CML 2 baseline method.

The data produced of the environmental impacts of the samples, as seen in

Table 4. Environmental impacts of each of the RCA concrete samples at Curtin University.

Impact Factor	Unit of Impact	M0	M1	M2	M3	M4
Land use	Ha a	0.002264	0.0022572	0.0022503	0.0022434	0.0022
Eutrophication	kg PO4eq	0.2340712	0.2340596	0.234048	0.2340363	0.2340
Water use	m3 H2O	4.3042941	4.1451434	3.9859926	3.8268419	3.6677
Global warming	kg CO2 eq	394.00318	393.76496	393.52674	393.28852	393.0503
Human toxicity	kg 1,4-DB eq	4544.3392	4547.3259	4550.3126	4553.2993	4556.286
Terrestrial ecotoxicity	kg 1,4-DB eq	1.18	1.17 × 10−11	1.17 × 10−11	1.16 × 10−11	1.156 × 10−11
Fresh water aquatic ecotoxicity	kg 1,4-DB eq	1.02 × 10−10	1.02 × 10−10	1.02 × 10−10	1.01 × 10−10	1.010 × 10−10
Marine aquatic ecotoxicity	kg 1,4-DB eq	5.43 × 10−8	5.47 × 10−8	5.52 × 10−8	5.57 × 10−8	5.615 × 10−8
Abiotic depletion	kg Sb eq	0.0002375	0.0002421	0.0002467	0.0002513	0.000256
Terrestrial acidification	kg SO2 eq	1.8304492	1.8333232	1.8361973	1.8390713	1.8419454
Ozone depletion	kg CFC-11 eq	6.98 × 10−6	7.13 × 10−6	7.27 × 10−6	7.42 × 10−6	7.56 × 10−6
Photochemical oxidant formation	kg NMVOC	1.8160063	1.8197836	1.8235608	1.8273381	1.8311154
Ionising radiation	kBq U235 eq	2.9716248	2.9709428	2.9702609	2.9695789	2.968897
Respiratory effects	kg PM2.5 eq	0.074594	0.0747299	0.0748659	0.0750018	0.0751378

Ha a—Hectare year, eq—equivalent, PO₄—Phosphate, H₂O—Water, CO₂—Carbon dioxide, 1,4-DB—1,4 Dichlorobenzene, Sb—Antimony, SO₂—Sulphur dioxide, CFC-11—Chlorofluorocarbon, NMVOC—Non-methane volatile organic compounds, U235—Uranium, PM2.5—Particulate matter (2.5 micrometres).

, shows the value of each of the impact factors associated with each individual unit of the impact. As the concrete samples are all very similar in composition, with identical binding chemicals, the differences between individual impact factors for each concrete sample are small. To compare each of the samples in a more direct way, the data collected can be normalised against the environmental impact of an average Australian person, as seen in Equation (1).

The normalised data for each of the M0–M4 concrete samples can be seen in Figure 3. Each unit of impact from each of the samples, as seen in Table 4 has been divided by the gross domestic environmental impact for the average Australian inhabitant, as seen in Table 1. The equivalent impact per inhabitant for each of the units of impact can be further multiplied by the weighting for each specific impact, as applicable to Australian conditions [16]. This gives a final unit for each of the concrete samples as an equivalency to an Australian inhabitant. The dominant environmental impact across all the concrete samples is human toxicity. This would be unusual for concrete made from ordinary Portland cement, where the highest weighted impact factor would usually be global warming emissions [20]. In this case, the concrete samples do not contain ordinary Portland cement, and instead contain an alternative binder of fly ash, sodium silicate and sodium hydroxide. Each of these individual components is known to be associated with human toxicity [21], so the results shown are in line with what would be expected. As each of the samples used the same chemical binding agents, with identical quantities, the only variation in environmental impacts between the samples will be due to the replacement of virgin aggregate with recycled glass aggregate.

3.4. Life Cycle Cost (LCC)

To produce an eco-efficiency portfolio comparing the different concrete samples, economic information must be determined for each of the samples to be used alongside the environmental impacts. The costs associated with the production of the samples arise from the constituent ingredients, transport and energy associated with the production of the samples. The components of the samples are fly ash, sand, natural aggregate, glass aggregate, alkali solution, electricity consumption and transport (by sea and road). The LCI in Table 3 was used to determine the associated cost of materials, transport and electricity for each of the concrete samples.

The prices of constituent ingredients were estimated from material prices specific to other concrete samples produced at Curtin University used for a separate sustainability assessment of recycled concrete mixes containing industrial by-products in 2019 [12]. The constituent prices in the study were obtained from Mr Mick Ellis, the Senior Technical Officer for Curtin Civil Engineering Laboratory. The material prices were based on large quantities of materials supplied to Curtin University between 2012 and 2018, so the prices are assumed to be similar to the samples M0–M4 produced for this LCC. The prices as shown in

Table 5. Estimates for material prices of concrete samples M0–M4.

	Fly Ash	Sand	NA 10 mm	NA 20 mm	GA 13–19 mm	GA 7–13 mm	Sodium Silicate	Sodium Hydroxide	Material Price
Price per ton	126.48	40.21	65.56	61.05	35	35	3600	5450
M0	400	684	592	592	0	0	27.08	10.83	309.56
M1	400	684	532.8	532.8	59.2	59.2	27.08	10.83	306.21
M2	400	684	473.6	473.6	118.4	118.4	27.08	10.83	302.85
M3	400	684	414.4	414.4	177.6	177.6	27.08	10.83	299.51
M4	400	684	355.2	355.2	236.8	236.8	27.08	10.83	296.16

are adjusted from initial values of fly ash AUD 120/ton, sand AUD 36.72/ton, natural aggregate 10 mm AUD 55.47/ton, and natural aggregate 20 mm AUD 55.75/ton.

The price for recycled glass aggregate was estimated to be AUD 35/ton. This is lower than the price paid by Curtin University for the actual recycled glass aggregate used in the M0–M4 samples, which was AUD 4500 for 250 kg. This estimate is based on the estimates by GGR Technologies Pty Ltd. which assume that if there were a market for recycled glass aggregates in Western Australia of 60,000 tonnes per year, the price for recycled glass aggregate would be reduced to AUD 35/ton [22]. The prices for sodium silicate and sodium hydroxide were estimated to be AUD 3.60/kg and AUD 5.45/kg, respectively, based on commercially available prices for Western Australia markets [23,24]

The prices for transport were estimated to be AUD 0.09/t-km for road freight and AUD 0.03/t-km for shipping, as described by the Australian Government Department of Infrastructure and Regional Development [25], estimated for 2021 when the concrete samples were produced at Curtin. The price for electricity during manufacturing was estimated from publicly available published information on the Synergy website, with the assumption that the tariff is a synergy business plan, priced at 37.5529 cents per unit (1 kWh) [26].

As some of the materials have been purchased in previous years, the prices paid by Curtin University must be adjusted for inflation up to 2021, when the samples were produced. The fly ash was purchased in 2019. Data from the Reserve Bank of Australia shows that the interest rate from 2019 to 2021 experienced a total change of 5.4%, resulting in an equivalent fly ash price of AUD 126.48 [27]. Similar calculations for sand and natural aggregate have adjusted the effective prices, as shown in Table 5.

The final estimated cost of each of the M0–M4 concrete samples can be seen in

Table 6. Total estimated cost of 1 m³ of concrete of each M0–M4 sample.

Total Cost (AUD)
M0	405.92
M1	402.91
M2	399.89
M3	396.89
M4	393.88

, which is the total cost inclusive of the material prices (seen in Table 5), the transport prices (as per the LCI travel distances) and the electricity prices (as per the LCI energy requirements), as detailed in the LCI. The estimated cost does not include any labour costs involved in creating the samples. Future studies may work to include the gate-to-grave component of the LCA.

3.5. Normalising Data Values to Produce the Eco-Efficiency Portfolio

The specific data relevant to the concrete samples M0–M4 is data for Australian total GDP, the population of Australia and the data on the strength characteristics of the concrete samples. This value is relevant in the application of concrete in high-strength applications, as concrete mixtures that have lower overall strength per unit of volume would require a larger volume of concrete to achieve similar required strength values, leading to increased costs.

The produce normalised economic values, data are obtained for the average GDP per inhabitant of an area, in this case Australia. The total GDP in the 2021–2022 financial year was AUD 2,090,347 million dollars [28]. The population of Australia in the middle of this period (31 December 2022) was 25.766 million [28]. This gives a GDP per inhabitant of AUD 81,128 in Australia. This figure will be used to determine the economic cost of each of the concrete samples, in terms of inhabitants per year of impact. Specific example calculations have been shown for sample M0 below. The normalised cost (Equation (1)) was determined using the cost for sample M0 as seen in Table 6. The normalised environmental value for global warming potential (GWP) for sample M0 is seen in Equation (2), where 394 kg of CO₂ is produced in the production of 1 m³ of M0, and the gross domestic production of CO₂ per inhabitant is 28,690 kg per year.

$$N_c=\frac{C{S}_{M0}}{GDP}=\frac{405.92}{81.128}=5.003\times {10}^{-3} ,$$

(9)

$$N_e{}_{M0, GWP}=\frac{EI}{GI}=\frac{394}{28.690}=13.73\times {10}^{-3} .$$

(10)

The normalised environmental value for each of the environmental impact factors can be summated, as seen in Equation (11).

$$E_e{}_{M0}={\displaystyle \sum }{N}_e\times W_i=0.04286095.$$

(11)

The initial portfolio positions (iPP) can be determined from the specific data produced for sample M0 alongside the averaged data for each of the samples M0–M4, as seen in Equations (12) and (13).

$$iP{P}_{M0, e}=\frac{E_e{}_{M0}}{E_e{}_{AV}}=\frac{0.04286095}{0.04289659}=0.999169,$$

(12)

$$iP{P}_{M0, c}=\frac{N_c{}_{M0}}{N_c{}_{AV}}=\frac{0.004829}{0.004778}=1.01083.$$

(13)

The initial portfolio positions must be further adjusted to effectively link the samples together. If the financial costs or environmental impacts of one option change, the other options may be better or worse choices relative to the changed option. Examples for Equations (14) and (15) are specific to the universal R value, and M0, respectively.

$$R=\frac{E_e{}_{AV}}{N_c{}_{AV}}=\frac{0.04289659}{0.004778}=8.980,$$

(14)

$$P{P}_{M0, e}=\frac{iP{P}_{AV, e}+\left(iP{P}_{M0, e}-iP{P}_{AV, e}\right)\ast \sqrt{R}}{iP{P}_{AV, e}}=\frac{1+\left(0.999169-1\right)\ast \sqrt{8.98}}{1}.$$

(15)

3.6. Eco-Efficiency Portfolio

Each of the five concrete samples M0–M4 were assessed to be plotted on an eco-efficiency portfolio. The PPMX, e and PPMX, c values were determined from normalised environmental and economic data as determined from the LCI in Table 3. Samples M0 (0% glass aggregate (GA)) and M1 (10% GA) were found not to be eco-efficient. This is due to the eco-efficiency portfolio determining the eco-efficiency of a sample, relative to the other options available. The comparison of samples MO to M1, relative to samples M2–4, indicates different results for the increased cost to produce the samples. The higher cost of samples M0 and M1 is due to the decreased percentage of relatively cheaper glass aggregate, as compared to the cost of the natural aggregate. Sample M2 (20% GA) was found to have a PPMX, e/PPMX, c ratio of one, making it eco-efficient. Samples M3 (30% GA) and M4 (40% GA) were the best performing samples, both being considered high in eco-efficiency. This is due to the cheaper cost to produce the samples, owing to the higher concentration of relatively cheaper glass aggregate making up a higher percentage of the sample. The eco-efficiency portfolio of the results can be seen in Figure 4.

The final portfolio positions of the samples are shown to be close to each other, with each consecutive sample being further to the right as compared to the previous sample. This can be explained by each of the samples having very similar environmental impacts through the production of the concrete. The environmental benefits of the replacement of virgin aggregate with recycled glass represent a minor portion of the overall environmental impact of the concrete samples. The incremental movement of each of the portfolio positions can be explained by the reduction in cost for each of the concrete samples, because the cost of recycled glass aggregate is cheaper than the cost of the virgin aggregate that is being replaced.

The samples that are not considered to be eco-efficient would be poor choices for concrete, as the ratio of economic benefits as compared to the environmental impacts is poor. The best concrete mixture for an eco-efficient solution would be the M4 mixture followed by M3 then M2 mixtures.

3.7. Hotspot Analysis

A hotspot analysis is a tool that can be used to evaluate the environmental impacts for each stage of the production process of a product or service. The SimaPro data analysis conducted during this research produced a hotspot analysis for each of the concrete samples M0–M4, as seen in Figure 5 and Figure 6. The hotpot analysis is a graphical representation of the energy consumption and CO₂ production associated with each of the three separate stages of production of the concrete samples. For each of the concrete samples, the left line represents the mining component of the production of raw materials, the middle line represents the transport of the materials to the final concrete mixing location and the right-side line represents the associated construction inputs; in this case, it is inclusive of the crushing energy associated with producing the recycled glass aggregate.

The hotspot analysis shows that for each of the samples the dominant pathway for the consumption of energy and to produce emissions is during the mining/production of raw materials, particularly for the alternative binder used in place of traditional Portland cement (the alkali solution). These results are consistent with idea that the reduction in environmental impacts due to the substitution of virgin aggregate with recycled glass aggregate constitutes a minor change in the total emissions produced during the production of concrete. The hotspot analysis shows the unique characteristic of geopolymer concrete, as compared to ordinary Portland cement concrete. When producing ordinary concrete, a hotspot analysis will identify the production of the cement as a major contributor to the global warming potential impact category, i.e., environmental impacts, with GWP being the dominant impact overall. As the geopolymer concrete does not require cement, it does not follow the same environmental impact pattern of GWP. Unfortunately, using fly ash alongside other alkali chemicals (sodium silicate and sodium hydroxide) presents the alternative challenge of an ecotoxic dominant environmental impact.

3.8. Monte Carlo Analysis

The modelling produced with the use of SimaPro for life cycle assessment analysis of concrete samples M0–M4 relies on many mathematical assumptions and probabilities. Large collections of inventory data are uncertain, usually falling within a range of values instead of specific, set values. An example of this can be seen in the environmental impacts associated with the production of concrete samples. Assumptions regarding the production of pollution associated with each unit of concrete have uncertainties. Different mining and transport methods will ultimately change the amount of pollution created for each scenario. To account for statistical variations due to uncertain data, SimaPro can run a Monte Carlo simulation. The Monte Carlo simulation involves running many iterations of the same calculations, with uncertainty distributions being taken into consideration, and randomised for each iteration. The benefit of this style of simulation is that random fluctuations and probabilities can be included to create a model that is likely to represent real world conditions.

For each of the concrete samples M1 and M4, a Monte Carlo simulation was performed in SimaPro. The simulation involved 1000 iterations at a 95% confidence interval. The impact factor of human toxicity is shown for M1 and M4 in Figure 7 and Figure 8, respectively. The data distribution shows a normal distribution of potential probabilities of the total equivalent amount of 1,4 Dichlorobenzene that would be produced due to the LCA parameters determined to produce the concrete samples. Human toxicity was chosen for this simulation as it was shown to be a dominant impact factor in the LCA for the concrete samples M0–M4. The coefficients of variance in both cases are around 9%, which is considered sufficiently reliable for most purposes by the Australian Bureau of Statistics [29].

The Monte Carlo simulation demonstrates the accuracy of the SimaPro modelling software. When using SimaPro LCA software databases, assumptions must be controlled, as creating models that are too complex needlessly increases the difficulty of creating reliable and accurate databases. Conversely, models that do not consider real world variations and simplify models in the pursuit of simplicity risk reducing the overall accuracy of the models. The results of the Monte Carlo simulation show that similar results are achieved using the Australian Indicator set database, as well as randomised iterations of outcomes based on the parameters of real-world variability.

4. Discussion

The ongoing development of new, innovative processes and techniques in the pursuit of sustainability is becoming more important in a world of growing population, decreasing natural resources and increasing awareness of the risks associated with unsustainable development practices.

By understanding the fundamental purpose of sustainable development, balancing the financial needs of businesses, the environmental limitations of the earth and the social requirements of society, there is potential to have ongoing, intergenerational and intragenerational sustainability. Using the eco-efficiency framework as a guide to assess new ideas for sustainable development is a recent application of sustainable engineering techniques. Eco-efficiency is about understanding that if non-renewable resources must be consumed by society, there should be efforts made to obtain the highest possible benefit from the smallest amount of resource use. Eco-efficiency is the culmination of an ongoing effort to grow and develop new ideas in the field of sustainable engineering.

The goal and scope of this study is suitable in the context of the glass aggregate concrete samples being prepared for the purpose of material characteristic and strength testing. The results of this research suggest there are potential eco-efficient benefits to using recycled glass aggregate as an aggregate replacement. There is potential for ongoing research to expand the boundaries of future studies to include the “gate-to-grave” component of the life cycle of concrete. The challenge of conducting studies with increased scope boundaries is the typical anticipated life expectancy of concrete structures, typically multiple decades. Ideally, concrete made from recycled glass aggregate would effectively perform the same function for the same time and produce the same end-of-life challenges and emissions as traditional, cement-based concrete. Future studies should work to determine if this is the case, as the alternative could be a concrete that is less eco-efficient than traditional concrete.

The life cycle inventory of this study can be considered sufficiently accurate for the purpose of assessing the viability of recycled glass aggregate concrete, but changes must be made in future research for large-scale production of glass aggregate concrete. The materials obtained and used in the concrete samples M0–M4 are scalable to larger amounts; however, the transportation distances are highly specific to the location of Curtin University. Future research should determine the best transportation methods and exact distances and location to be used at a commercial scale. Particular attention needs to be focused on the collection and transportation of the recycled material.

The SimaPro software and associated databases have been demonstrated to be capable of effectively determining the environmental impacts associated with the individual concrete samples, as described by the life cycle inventory. Environmental impacts specific to Australian conditions were determined using multiple SimaPro databases. By using known Australian environmental impacts, with existing advised weighting factors, data could be normalised into standardised units. Future studies have the potential to use more specific Australian environmental impacts, derived from Australian databases, in place of European specific databases (eg ReCiPe midpoint). Alternatively, different normalisation factors could be determined to match the data available for Australian specific databases, for more accurate Australian data.

The change in environmental impacts between the concrete samples, as calculated by SimaPro, shows the incremental change in environmental impacts resulting from a decrease in use of quarried, virgin aggregate and increasing use of recycled glass aggregate, while other factors remain constant throughout the samples. Notably, land usage and water usage decrease as glass content increases in the samples, and global warming potential and human toxicity both increase as the glass content increases. This is likely due to the scale of the Spartel PTY LTD glass crusher being used [The glass crusher utilised is effectively limited to being fed a single glass bottle at a time, making it inefficient in comparison to larger crushers that would be used if commercial quantities of glass were being produced at scale].

Life cycle costing data were obtained for material pricing specific to materials at Curtin University used in the production of the concrete samples. Inaccuracies due to the historic purchase of the material by Curtin have been rectified by adjusting prices for inflation, but future studies should attempt to determine current market pricing to make a more accurate reflection of ongoing market conditions that may make present prices different to adjusted historic pricing. Ongoing global events such as fuel shortages and insufficient global shipping supply in the time between the materials being delivered to Curtin have potential to reduce the accuracy of the financial data used in this report.

The normalisation of data from fourteen individual environmental impacts to a single “per inhabitant” unit is derived from existing literature. The method appears to effectively convert multiple, weighted variables into a unit that can be easily contrasted with cost using an eco-efficiency portfolio. The graphical representation of previously tabulated data is an efficient communication method. The eco-efficiency portfolio effectively compares each of the M0–M4 concrete samples. The addition of a traditional, cement-based concrete sample would be beneficial to contrast the perceived benefits (or lack thereof) of glass aggregate geopolymer concrete against traditional concrete.

There are potential advantages of sourcing concrete additives that would otherwise be harmful if disposed in a landfill, if the result was that binding into concrete reduced potential leaching into the environment-. Phthalates are a group of chemicals used in the production of many flexible polyvinyl chloride plastics such as garden hoses, medical tubing and raincoats [30]. If disposed of in landfill sites, phthalates are known to leach into the environment [31], causing long lasting or permeant damage to the ecosystem. By applying eco-efficiency principles, the goal is to achieve the greatest benefit with the fewest consequences to the environment and future generations. In a situation where a limited amount of plastic can be added into concrete, the benefits can come in the form of improved concrete characteristics or improved environmental consequences as a result of deferring phthalate-containing plastics from landfill for use in concrete. Further research should be conducted to determine the potential environmental benefits of deferring the most harmful landfill waste into concrete additives, the potential material property consequences such waste would have on concrete and the logistical challenges of deferring only the most harmful waste from landfill.

5. Conclusions

The results from the eco-efficiency portfolio demonstrate the benefits of the application of eco-efficiency studies in the pursuit of sustainability. The fundamental idea behind the use of recycled glass aggregate in concrete as a virgin aggregate replacement is to decrease the environmental impacts resulting from quarrying of virgin materials. Although recycled aggregate was used in the concrete mixes to replace the energy intensive virgin aggregates, mining to material production still accounts for more than 70% of the total GHG emissions. This study has demonstrated that although the reduction in environmental impacts is relatively small when using recycled glass aggregate, there is an additional economic benefit from the reduction in cost of materials contributing to the concrete. Accordingly concrete mixes using recycled glass between 20% and 40% have been found to be ‘eco-efficient’. Concrete mixes using 40% recycled glass aggregates are about 7% and 12% more efficient than the mixes using 30% and 20% recycled glass aggregates, respectively. This reduction in cost helps to improve the overall eco-efficiency by achieving more through the same use of resources. This is the fundamental principle of eco-efficiency: achieving more with the resources you use.