Innovating traditional building materials in Chembe, Malawi: assessing post-consumer waste glass and burnt clay bricks for performance and circularity

ABSTRACT

Across the Global South, post-consumer waste glass is an often dumped, and under-utilised resource. Even in Malawi, with widespread return schemes, many barriers exist, inhibiting reuse, and necessitating appropriate solutions. The purpose of this article is to evaluate the performance of post-consumer waste glass as a coarse aggregate within burnt clay bricks, and to assess the feasibility for the recovery of this waste material from dumped stocks within Chembe. Using a brick design and testing methodology, which could be replicated within a rural African context, we tested a range of glass additions (both quantity and size of particle) for compressive strength and water absorption properties. Our results suggest that waste glass can function as a performance enhancer, with positive effects on compressive strength observed at up to 10% crushed waste glass content. These findings support existing literature on glass waste additions, yet show that optimal results can also be had with post-consumer waste glass and in low income, and less technology-reliant contexts. Moreover, our findings suggest that current above-ground stocks of waste glass are sufficient to support the production of hybrid building materials for decades, however further innovation is necessary in order to achieve a sustainable mode of practice.

KEYWORDS:

• Landfill Mining
• solid Waste
• Africa
• circular Economy
• alternative Building Materials

1. Introduction

In a global climate of diminishing natural resources and ever-increasing stockpiles of solid waste, landfill mining has emerged within waste management studies as an ever-growing topic of academic and practical discourse. Concerned with shifting patterns of consumption from limited, underground, stocks of resources, to the above-ground materials we hoard away in landfills, dumpsites, tailing ponds, and other waste management spaces, landfill mining aims for the recovery, circular use, and recycling, of these anthropogenic materials (c.f. (Baccini and Brunner 2012; Cossu and Williams 2015; Arora et al. 2017; Kumar and Evelyn 2019)). Yet, what potential does the concept hold for Africa: a continent with few engineered landfills, and severe inequalities in waste management service provision? Moreover, although the topic has produced substantive literature over the past decade, the vast majority of landfill mining pieces to emerge in both theory and practice have been set within the Global North, and have been designed to accommodate Western settings, cultures, and institutions. Moreover, the few narratives of circularity, or even urban mining, that have been contextualised within African settings have narrowly focused on informality (see, for instance, Grant and Oteng-Ababio 2016), as if to minimise the very legitimacy or sustainability of any possible innovations. This in itself, is clearly problematic, and speaks in degrees to the ways in which the Global North views African cities. Yet, what about beyond Africa’s cities – to its burgeoning peri-urban areas or densifying villages? In African nations, these are also often spaces of rapid growth, yet they generally suffer from poorer public services, including access to waste management services, as well as fewer opportunities for employment or economic development (c.f. (Kombe 2005; Thornton 2008; Doan and Oduro 2012; Kalina 2020; Kalina and Tilley 2020; Kleemann et al. 2017; Nuhu 2019; Tilley and Kalina 2020). What relevance do landfill mining principles have in a context for which a landfill, or even basic waste management services, is a distant impossibility? How can these spaces, as growing centres of consumption and waste, contribute to meeting society’s ever-increasing need for usable raw materials?

Malawi is consistently ranked among the poorest, least developed, and most corrupt countries in the world. Despite its challenges, it boasts diverse, unspoilt, natural beauty, a peaceful society, and limitless tourism potential. Spanning a narrow strip of sand and clay no more than 3 km long and less than half a kilometre deep, and sandwiched between green, thickly forested hills, and the transparent blue waters of Lake Malawi, Chembe is perhaps Malawi’s most famous village. As recently as the late 1990s Chembe remained a sleepy fishing village of around 2,000 people, with basic tourism offerings and only a handful of lodges (Smith 1993). Over the past two decades the population has exploded, however, with improvements in social services dramatically increasing the life expectancy of residents, and consistent growth in the tourism industry offering sustainable income-earning opportunities, encouraging young adults that may have once migrated to the city to stay and work, while drawing migrants from other villages from the region. As a result, the once sleepy fishing village of 2,000 now has nearly 20,000 individuals, and continues to grow.¹ Despite the extension of services such as electricity and water to the village, Chembe’s waste management services have not kept pace with its growth. Moreover, growth in household consumption and the trebling of the population has dramatically increased the amount of waste produced by individual households, while the increase tourists and tourism infrastructure has dramatically increased the amount of single-use products being discarded within the village (c.f. (Kalina et al. 2021). As a consequence of these factors, the detritus of Chembe’s prosperity lies in heaps throughout the village, strewn on the side of paths, pushed into piles, or collecting in low-lying areas that have become unofficial dumping grounds.

Of the rubbish littering Chembe’s pathways, the omnipresent shards of glass waste are both the most noticeable and the most dangerous. With no waste management services, these bottles generally find their way to the many piles described previously, and once broken, they present persistent environmental safety risks to residents. A rough quantification, conducted as part of a GIZ survey, estimates that tourism lodges within Chembe, alone produce nearly 200 kg of glass waste a week (Augustine 2019). Calculating back for 20 years of increased consumption, it was estimated that the lodges have produced, and dumped, more than 200,000 kg of glass waste over that period. Again, although these are major sources, these estimates do not include other sources of glass waste, such as restaurants, bars, and local household consumption, so the true amount of glass waste dumped throughout the community is likely significantly higher. The efforts of some of the lodges participating in the collection scheme has succeeded in consolidating some of the glass waste; in fact their ad hoc dump mostly consists of heaps of broken glass (Figure 1), however, with no solution for disposal it is a problem that continues to grow, as the glass waste keeps piling up.

The purpose of this article is to evaluate the performance of post-consumer waste glass as a coarse aggregate within burnt clay bricks, a popular and widely available domestic building material within Chembe, and Malawi more broadly. The investigation also assesses the feasibility for the recovery of this waste material, from dumped stocks within Chembe, to contribute to local sustainability and the circularity of resource use within the community. Our results follow with other investigations into the use of waste glass within burnt clay bricks, which suggest that waste glass can function as a performance enhancer, with positive effects on compressive strength observed at up to 10% crushed waste glass content. However, in contrast with past literature, which has largely emphasised the reuse of either cleaned and sorted post-consumer waste glass or post-industrial waste glass within our settings our intention was to develop a methodology and product that could be easily replicated in Chembe, with minimal resources. Our findings suggest that significant potential for `mining’ practices exists within Chembe, and that current above-ground stocks of waste glass are sufficient to support the production of hybrid building materials for decades. However, other aspects of the brick-making process, such as the firing, challenge its long-term sustainability, necessitating further innovation in technology and practice.

2. Landfill mining, waste and bricks

The purpose of this brief literature survey is twofold. First, it highlights recent developments within waste management studies discourse, which have attempted to reframe landfill mining concepts and practices within Global South contexts. These contributions, particularly those from the Indian sub-continent, have reconceptualised landfill mining as `dumpsite mining’, adapting the principles of landfill mining that work in their context and disregarding those that do not, while allowing greater space for informality. Second, it evaluates current literature on the utilisation of waste glass in traditional building materials, specifically burnt clay bricks, while acknowledging the potential that remains for innovation within this specific material framework.

2.1. Landfill mining in the global south

Landfill mining is one of the many concepts that have emerged within waste management discourse to refer to the recovery, recycling and extraction of resources from waste. Specifically, landfill mining represents the activities involved in extracting and processing wastes which have been previously stocked in particular kinds of deposits (municipal landfills, tailing ponds, etc.) (Cossu and Williams 2015). In this regard, it is distinct from urban mining which looks to reclaim valuable elements from any type of anthropogenic stocks, such as industrial sites, buildings, infrastructure, etc. (Koutamanis, van Reijn, and van Bueren 2018). Like most material recovery schemes, the suitability of landfill mining is usually rooted in its economic feasibility, and, as such, it has historically been utilised to recover high-value elements (rare metals from waste electrical and electronic equipment (WEEE) or phosphorous from sewage ash, for instance). However, as Cossu and Williams (2015) note, there is significant scope for recovery from a broader range of waste streams than are currently utilised. Nonetheless, landfill mining (and most material recovery and circular economy strategies, for that matter), have largely been developed within the Global North for Global North waste management and socio-economic systems.

Studies on landfill mining have been historically concerned with the use of highly technical processes to recover valuable resources from engineered landfills, predominantly in the Global North. However, the past decade has seen a broadening within the landfill mining literature to include a number of applications within Global South contexts, using landfill mining principles, but often more technologically basic processes, to reclaim and recover a variety of recyclable materials from old, often open, un-engineered dump sites. These works have largely emerged from the Indian sub-continent and Southeast Asia, and emphasise the reclamation of abandoned municipal solid waste (MSW) dumps for recoverable materials, primarily metals and rare elements, but also soil-like materials that can be utilised in construction and infrastructural applications (see (Prechthai, Padmasri, and Visvanathan 2008; Somani et al. 2018; Mohan and Joseph 2020; Somani et al. 2020). A commonality across these applications is an emphasis on minimising the processing and treatment of waste fractions prior to use in order to minimise cost and required technological inputs. Finally, Mutafela et al. (2018) specifically examine the feasibility of mining abandoned glass dumping sites in Sweden. Their investigation is primarily concerned with using GPS technology to map abandoned glass dumping sites, in the absence substantive initial site details. Their results suggest that glass caches, which contain high glass content, could be precisely excavated with a need for minimal post-excavation sorting (Mutafela et al. 2018).

2.2. Waste glass and building materials

The concept of using glass as an additive to building materials is not new. In the North, emphasis for innovation has centred on foam glass, made from either molten glass or sintered glass particles, with some products utilising up to 98% post-consumer waste glass content (El-Haggar 2007). However, these processes require immense energy inputs and initial investments, and have seen little adaptation within the Global South. Investigations into the feasibility of integrating waste glass into burnt or fired clay bricks were carried out as early as 1970, but interest has renewed within the past two decades, particularly within Global South contexts (Matteucci, Dondi, and Guarini 2002; Chidiac and Federico 2007). Investigations have centred on two contributing properties of waste glass (WG), which make it suitable is an additive to bricks, namely water absorption and compressive strength. However, most investigations have utilised post-industrial glass waste or sorted post-consumer glass waste for their experiments (Chidiac and Federico 2007), unlike our investigation which centres on unsorted post-consumer glass waste. The sorting and cleaning of glass diverted from MSW for re-use in bottle industries is known to be time-consuming and costly, with bottles often arriving contaminated with other wastes (Achintha 2016). Cutting these additional steps adds to the sustainability of the items reuse and circularity.

Waste glass and natural sand have been found to have similar physical properties, however WG has a lower water absorption rate than sand, by about 14% (Jani and Hogland 2014). Investigating the use of WG in concrete, Ismail and Al-Hashmi (2008) found that concrete with WG as an aggregate exhibits lower water absorption rates (a lower water absorption rate will contribute to increase properties of concrete and elevate the quality closer to ‘extremely good concrete’ where levels of water aborbtion rate are close to 5–6% (Khatib and Mangat 1995; Neville 1995). As an additive to bricks, glass acts as a flux owing to its Na₂O content and its non-crystalline composition. Moreover, it contributes to sintering processes at lower temperatures. The results exhibited in most previous research indicated an increase in compressive strength with increasing waste glass contents, especially between 10% and 30% mass (Tucci et al. 2004; Chidiac and Federico 2007; Lin 2007; Demir 2009; Loryuenyong et al. 2009; Phonphuak, Kanyakam, and Chindaprasirt 2016; Shihada 2017). Fired clay bricks with appropriate mechanical properties can be attained (given suitable firing temperatures) when using waste glass contents ranging from 15 to 30% (by weight of clay) (Loryuenyong et al. 2009; Shihada 2017). Moreover, Chidiac and Federico (2007) showed that the strength and transport properties of clay bricks were enhanced due to an improvement in pore structure at a clay addition of 15% (by weight), for both fine and coarse waste glass. Finally, Phonphuak, Kanyakam, and Chindaprasirt (2016) and Shihada (2017) have suggested that that the use of 10% waste glass and firing at 900°C yielded bricks with strengths similar to those of normal clay brick fired at 1000°C. This finding points to the value of using waste glass additives in low-income contexts, were high or consistent firing temperatures may be more difficult to obtain.

3. Materials and methods

3.1. Quantifying Chembe’s waste glass stocks

A simplified waste quantification based specifically on the glass waste generated by tourists within Cape Maclear quantified the approximate amount of glass waste that is currently stored within local glass dumpsites. The glass waste generated in the village by residents was not assessed because there are two distinct glass waste streams in Chembe: bottles for beverages produced within the country which carry refundable deposits, and second, imported drinks, which are outside the deposit scheme.

Malawi has high rates of glass bottle reuse, as a number of its major brands (Coca-Cola, Carlsberg Beer, Fanta, Sprite, and a local Malawian beverage named Sobo) carry deposits on their bottles. These bottles are generally collected and returned, and few make it to a landfill or are discarded. However, current schemes do not include all glass bottle waste in the country, and a number of major brands, particularly imported brands (such as Heineken and drinks made by South African Breweries) do not participate. As a result, these bottles are imported into the country and dumped, with no accountability for their reuse or disposal. In most contexts, this is not a problem, as imported beverages are much more expensive than the locally produced brands; the average Malawian village consumes very little Heineken. However, in bustling Chembe, every lodge has a bar, and tourists are less likely to notice the price difference, i.e. the South African cider ‘Savanna’ being is one of the most popularly consumed drinks by international guests. Thus, overwhelmingly Chembe’s glass stocks can be attributed to tourism, with local consumption making little contribution, as those bottles are predominantly returned for deposit. Very few refundable bottles make it to Chembe’s dumpsites, and even a quick visual survey will show the overwhelming preponderance of imported brands within the stocks (see Figure 1).

Figure 1. Glass waste in Chembe

The quantification is based on the 20 lodges that run along the edge of Chembe. Lodge owners estimate generating an average of 50 bottles of waste glass per week, at an average mass of 9.85 kg. Moreover, interviews conducted with lodge owners and local officials suggest that glass stocks began to accumulate approximately 20 years ago, when tourism numbers began to grow: imported beverages became more available, and individualised waste management strategies began to coalesce into more centralised dumps as the community’s waste footprint grew.² At 9.85 kg of waste glass per week, per lodge, spread over this 20-year period, we estimate the amount of stored waste glass within Chembe as roughly 204,880 kg.

3.2. Traditional building methods in Chembe

Although Cape Maclear and Chembe are among the top destinations for tourists within Malawi, and the community is far more prosperous than the average Malawian village, traditional architecture continues to dominate, with everything from lodges, to shops, to homes being predominantly constructed from handcrafted burnt clay bricks, as they have been for generations. Some precast concrete blocks are also used, and cement is used for a variety of building applications, such as slabs, pillars, porches, and for mortar. However, due to its geographical isolation, cement is not readily available within the village, and is expensive. Moreover, clay is found locally in abundance; the flood plain that separates the village from the national park, and which is planted seasonally with maize, contains a seemingly inexhaustible supply, enough to supply the trade of several full time brick makers. Most importantly, however, for the residents of Chembe, burnt clay bricks remain aspirational. The lodges build in clay brick because they want to capture a traditional look, while locals also prefer the clay brick aesthetic to more modern precast concrete, and many new homes that could afford other materials, continue to use clay bricks for at least part of the construction. As noted, clay brick manufacturing is a specialised craft in Chembe, supporting several livelihoods.³ The clay for the bricks is excavated from low-lying areas within the flood plain behind the village, and then mixed with water and sand⁴ in order to obtain the right consistency. Bricks are then moulded using wood moulds, trimmed with a machete, and then set out to dry in the sun for several days (Figure 2). The brick makers consulted for this study described making two sizes of bricks: a small one (200 cm (length) x 100 cm (width) x 70 cm (height)), and a larger one (290 cm (length) x 130 cm (width) x 80 cm (height)). However, other sizes of bricks were observed within the village, and are presumably available.

Once dry, bricks are transported to the building site and stacked into an oven shape, where they are fired several times, using wood gathered (illegally) from the adjacent National Park (see Figure 2). During this period, the ‘oven’ structure is dissembled and reassembled a number of times so that the bricks are able to fire evenly. Once sufficiently fired, the bricks are sufficiently weather-proof, and can used as needed. The process of crafting the bricks from clay is highly streamlined, and an efficient brick maker can cast over 1,000 bricks per day. Firing the bricks, however, is extremely inefficient, as it relies on an increasingly scarce resource (wood), and remains an object for further investigation.

Figure 2. Bricks being fired in Chembe

3.3. Replication in the workshop

The following describes the processes used to re-produce clay bricks and waste glass materials within a laboratory setting with similar properties to those found within Chembe. All efforts were made to ensure consistency in materials and practices.

3.3.1. Preparation of waste glass

For our experiment, we utilised post-consumer waste glass bottles (330 ml beer bottles), of a brand commonly available within Chembe. In order to remove potential contaminants, all bottles were soaked for 15 minutes, and then scrubbed with a sponge and dishwashing soap to remove the labels. The bottles were then rinsed with clean water and left to dry in the sunlight for approximately 15 minutes.

Clean bottles were manually crushed with a sledgehammer inside of a steel drum. The crushed glass particles were passed through a 19 mm sieve; the particles that could not pass through the sieve were returned to the drum to be crushed further. Finally, the crushed and sieved WG was thoroughly mixed, and a sample of 500 g was taken to obtain a general distribution of the particle sizes.

Once prepared, a particle size distribution was determined per the South African Bureau of Standards (SABS) values for coarse aggregates and using SABS approved sieves. An indication of the typical particle sizes that are generated in a ‘typical crush’ (crushing the waste glass such that it passes the 19 mm sieve) is shown in (Table 1). These results served as a means of approximating the number of glass bottles that were required to obtain the desired quantities of the various sizes of WCG for the testing procedures

Table 1. General distribution of particle sizes of WCG in a ‘typical crush’

Sieve Size (mm)	Mass Retained on Sieve (g)	Percentage of Total (%)
19	0	0
9.5	211	42.2
4.75	106	21.2
Pan	183	36.6
Total	500	100

3.3.2. Clay mix design

The clay mix used in the experiment was designed utilising guidelines designated by Corobrik⁵ and had a moisture content of 18%. The Clay Brick Association of South-Africa recommends an acceptable water absorption for clay bricks to range between 12% and 20%: values below 12% may cause difficultiles in obtaining a proper bond between the mortar and the bricks (The Brick Industry Association 2006). Moisture content affects the compression strength of both bricks and cement mortar, while with bricks, the greater the moisture content, the lower compressive strength (Navaratnarajah and Rumeshkumar 2018). Moreover, the understanding of the effect resulting from the material microstructure and the hygrometric properties of clay bricks plays a fundamental role in controlling the condensation phenomena directly affected by high levels of moisture content, which in turn, can lead to the deterioration of the brick structure (Raimondo et al. 2007). The moisture content of the base clay mix was calculated with the aid of Equation 1(1) $MC=\frac{M_{wet}-M_{dry}}{M_{dry}}\times 100$(1) , where MC is the moisture content of the clay sample as a percentage, M_wet and M_dry are the masses of the clay material and WG in the wet state and dry state in grams, respectively.

(1) $MC=\frac{M_{wet}-M_{dry}}{M_{dry}}\times 100$(1)

As can be seen in (Table 2), the average moisture content of the clay material was 2.33%, which was used to offset the desired moisture content of the clay mix. The desired moisture content was 18% and thus the moisture content of the clay brick mix was increased by 15.67%.

Table 2. Determination of the moisture content of the clay material

Test Number	Mass of Wet Clay and Glass Bottle (g)	Mass of Dry Clay and Glass Bottle (g)	Moisture Content (%)
1	97	96	1.04%
2	85	82	3.66%
3	82	80	2.50%
Average	88	86	2.33%

Water was added to achieve the desired moisture content and, for the test samples, WG was added as a substitution by mass of the clay material. The proportions of water and clay were calculated by manipulating Equation 1(1) $MC=\frac{M_{wet}-M_{dry}}{M_{dry}}\times 100$(1) to solve for Mwet by setting the moisture content as 18 % and using Mdry as the mass of clay required less the water already in the material based on the moisture content of the ready mix clay material. The proportions of glass for the various substitutions were calculated accordingly. The water requirements for the clay mix are shown in (Table 3) the sample 0% WG being a Control Sample (CS).

Table 3. Calculation of water requirements for clay mix design

WCG Substitution (%)	Mass of 6 Bricks (kg)^6	Water Required (kg)	Mass of Glass Required (kg)	Total Mass of Bricks (kg)
0% WG (CS)	21.00	3.29	0.00	24.29
2% WG	20.58	3.22	0.42	24.22
5% WG	19.95	3.13	1.05	24.13
10% WG	18.90	2.96	2.10	23.96

In order to achieve a uniform brick size, a clay brick mould was made from sheet metal. The mould was designed with dimensions to suit that of an imperial brick (The Claybrick Association 2015). The imperial brick dimensions are; 222 mm (long) x 106 mm (wide) x 73 mm (high) and has a mass that varies between 2.4 kg and 3.3 kg depending on the clay properties (The Claybrick Association 2015). This size was maintained throughout the experiment to ensure uniformity in the testing procedure, and is comparable in size to the smaller clay brick typically produced in Chembe (200 cm x 100 cm x 70 cm).

3.4. Preparation of samples

A total of 42 bricks were produced in seven batches of six bricks each. The batches consisted of a 0% substitution (CS), 2% WG small, 5% WG small, 10% WG small, 2% WG large, 5% WG large and 10% WG large. The density of the clay material used for the calculation of the quantities in the mix design was used as 1759.2 kg/m³.

During the manufacturing process, the brick mould was cleaned and dipped in water to prevent the material from sticking to the mould. The mixture was filled in three layers of approximately equal depths and compacted between layers with a wooded compaction tool to remove air from the mixture, and the top of the brick was levelled using the tool. The clay brick was demoulded and marked with a steel marking tool.

The engineered bricks were left to air dry for three days, after which they were transported to Corobrik. At Corobrik were dried for a further 3 days. Following this second drying period they were fired in Corobrik’s kiln gas firing mode, which reaches a maximum temperature of 950 degrees Celsius for approximately 48 hours. Due to the difference of context between Corobrik facilities and the facilities and system available at Chembe, the effect of temperature applied was not considered for comparison purposes. Firing temperatures for bricks made at Chembe is assumed to be similar as those produced at Corobrik, however, likely less consistent. The impact of the different firing processes on the brick remains a space for further investigation.

3.5. Testing protocols

3.5.1. Compressive strength test

The compressive strength of the clay brick sample is given by Equation 2(2) ${\mathrm{f}}_{\mathrm{c}\mathrm{b}}=\frac{\mathrm{F}}{{\mathrm{A}}_{\mathrm{c}}}$(2) , where f_cb is the compressive strength of the burnt clay brick in mega-Pascal’s, F is the load applied at failure in Newton’s and A_c is the cross-sectional area on which the compressive load is applied in millimetres squared.

(2) ${\mathrm{f}}_{\mathrm{c}\mathrm{b}}=\frac{\mathrm{F}}{{\mathrm{A}}_{\mathrm{c}}}$(2)

The compressive strength of the clay bricks was tested using the IS 3495 Indian Standard (2002) with minor variations. Samples were placed in a compression strength test machine⁷ and axially loaded at a uniform rate until sample failure.

3.5.2. Water absorption test

Results were analysed per the IS 3495 Indian Standard (2002). The water absorption percentage is given Equation 3(3) $\mathrm{W}\mathrm{A}=\frac{{\mathrm{M}}_2-{\mathrm{M}}_1}{{\mathrm{M}}_1}\times 100$(3) , where WA is the water absorption as a percentage, M₂ is the mass of the specimen after it had been removed from the water bath in kilograms and M₁ is the mass of the specimen before it had been placed in the water bath in kilograms.

(3) $\mathrm{W}\mathrm{A}=\frac{{\mathrm{M}}_2-{\mathrm{M}}_1}{{\mathrm{M}}_1}\times 100$(3)

The samples were placed in a water bath for 24 hours. After, the samples were removed and weighed within 3 minutes of exiting the water bath, and their masses were recorded.

4. Results and discussion

For each sample set of 6 bricks, 3 were tested under compression loading and 3 were tested in terms of the water absorption characteristics, these results were averaged and the discussion presented is based on this average.

4.1. Compressive strength

The average compressive strengths of the burnt clay bricks are illustrated in (Table 4), where S-WG represents the waste crushed glass with a particle size defined as small (4.75–9.50 mm) and L-WG represents the waste crushed glass with a particle size defined as large (9.50–19.00 mm).

Table 4. Average compressive strength of the burnt clay bricks

Sample Identification	Average Compressive Strength (MPa)	Standard Deviation
0% WG (Control Sample)	11.788	0,212
2% S-WG	14.043	2,566
2% L-WG	14.339	0.398
5% S-WG	15.766	1.875
5% L-WG	13.849	0,499
10% S-WG	12.563	0.252
10% L-WG	13.535	0,230

Results suggest that a substitution of WG of up to 10% in a clay mixture increased the compressive strength of the burnt clay bricks. The greatest strength improvement was exhibited in the 5% S-WG sample which showed an increase of 3.978 MPa (33.75%) as compared to the control sample. At a substitution percentage of 2% WG the small and large WG performs in a similar manner with a slight variation in strength, the 2% S-WG and 2% L-WCG showed an increase in strength relative to the control sample of 19.13% and 21.64%, respectively, with the larger particles outperforming the smaller particles at this substitution percentage.

(Table 5) highlights the increase in compressive strength of the samples relative to the control sample, it also emphasises on the impact of the waste glass on the improvement of the compressive strength of the brick.

Table 5. Increase in compressive strength of burnt clay bricks with WG as compared to control

Sample Identification	Strength Increase Relative to Control (MPa)	Strength Increase Relative to Control (%)	Rank (Descending)
2% S-WG	2.255	19.13	3
2% L-WG	2.551	21.64	2
5% S-WG	3.978	33.75	1
5% L-WG	2.061	17.48	4
10% S-WG	0.775	6.57	6
10% L-WG	1.747	14.82	5

From (Table 5), it is possible to observe that at a substitution percentage of 5% WG the small and large WG both yield a clay brick with increased compressive strength as compared to the control sample, however, the small WG yields a clay brick with a substantially higher compressive strength as compared to the large WG with a difference in the percentage of 16.27%. It is evident that at a 5% WCG substitution the small particle size outperforms the large particle sized WG.

At a substitution percentage of 10% WG the large and small WG both yield a compressive strength that is greater than the control sample, however, the small particle WG had a compressive strength of 8.25% lower than the large particle WG as compared to the control sample. It is evident that the large particle at a 10% WG substitution outperforms the smaller particle WG. In fact, the 10% WG substitution was ranked at the bottom of the data set for compressive strength as compared to the control sample (see Table 5), suggesting that the optimum WG substitution value may lie below a substitution of 10%.

These findings are slightly at odds with the literature. For instance, Demir (2009) shows a linear increase in compressive strength up until a 10% WG substitution whereas our results clearly show a peak at a 5% WG substitution after which there is a decline in compressive strength. However, our observation of an increasing trend in compressive strength up until a 5% WCG substitution is reflective of Demir (2009) findings. The increase in percentage of aggregates at 10% revealed the existence of a threshold and confirmed Demir (2009) findings. We can assume that an increase of percentage above 10% would result in a continued decrease in compressive strength.

4.2. Water absorption

Results of the water absorption testing (see Table 6) showed the water absorption of the sample clay bricks increase marginally as compared to the control sample for the 2% S-WG, 2% L-WG and 5% S-WG by 0.127% (1% of the control), 0.196% (1.6% of the control) and 0.454% (3.6% of the control) respectively. Water absorption decreased for the 5% L-WG, 10% S-WG and 10% L-WG by 0.013% (0.1% of the control), 0.018% (0.14% of the control) and 0.071% (0.56% of the control).

Table 6. Average water absorption (%) of the burnt clay bricks in function of waste glass content (%)

Sample Identification	Average Water Absorption (%)
0% WG (Control Sample)	12.575
2% S-WG	12.704
2% L-WG	12.771
5% S-WG	13.029
5% L-WG	12.562
10% S-WG	12.557
10% L-WG	12.504

The noted decline in average water absorption percentages at higher concentrations of WG is supported by Demir (2009), however, in said study, the water absorption characteristic of the clay bricks was significantly higher as indicated in (Table 6). The increase in average water absorption percentages is however not supported by Demir’s findings. The average water absorption percentages are relatively constant with a maximum range of 0.525%, possibly attributable to the increase in voids surrounding the glass particles due to the inconsistent particle geometry of the WG.

The marginal variation is attributed to aggregate shape and size arrangement in the matrix. It is impossible to predict and organise the matrix and its aggregates, and while the mixture of the components is calculated precisely, parameters like water absorption will vary. However, the variation exhibited could be considered marginal in this particular case.

4.4. Viability of waste glass in clay bricks

In terms of average compressive strength, the bricks manufactured for this experiment performed better than the control sample. Using South-African National Standards (South African National Standards 2011) for building requirements, a number of practical applications become apparent (see Table 7).

Table 7. Suggested bricks with WG for various applications (South African National Standards 2011)

Possible Applications	Average Required Compressive Strength for Solid Units (MPa)	Proposed Brick Specification
Single Storey or the Upper Storey of a Double Storey Building	4.0	10% L-WG
Lower Storey of a Double Storey Building	10.0	10% L-WG
Infill Panels in Concrete and Steel Framed Buildings of four Storeys or less	4.0	10% L-WG
Free Standing, Retaining, Parapet and Balustrade Walls	5.0	10% L-WG

4.5. Waste glass mining and circularlity potential within Chembe

The viability of utilising waste glass within burnt clay bricks suggests the feasibility of a circular economy strategy to mine stored glass stocks within the community for reuse. We previously estimated these stocks to be roughly 204,880 kg. Our satisfactory results with a simple hand crushing method suggest that, for maximum waste utilisation, both small-sized and large-sized particles should be used. For our calculation we assume a concentration of WG of 10% because single-story housing would be the most common usage for the materials, and for this building application a 10% WG yields a sufficient strength (see Table 7). Finally, as described, the brick sizes in Chembe vary, however, the small bricks observed have similar dimensional properties to the bricks tested in this study and as an approximation, were applied in this analysis.

Based on the calculated estimate of 204,880 kg of waste glass dumped in Chembe, we calculate it would take 586 days to fully utilise the waste glass currently stored, at current brick production rates, resulting in 585,371.4 bricks, which utilise the 10% WG substitution (.35 kg of WG per brick). However, this result assumes, unrealistically, that there is no further accumulation of waste; this calculation addresses only the existing backlog of glass. Additional consumption, over time, would prolong the viability of the WG mining process.

5. Conclusions

Within the North, circular economy strategies, such as landfill mining, urban mining, and recycling, have contributed to sustainable resource recovery and reuse. However, within the Global South, where many contexts lack engineered landfills, or even consolidated dumpsites, what relevance does the concept hold? Within less technology-reliant contexts, low resourced spaces, how can resources be maximised to improve the environment and livelihood of local communities?

The purpose of this article was to assess the feasibility of one circularity strategy in Chembe, Malawi, utilising stockpiled and unsorted crushed post-consumer waste glass as a partial aggregate replacement in burnt or fired clay bricks, the most common building material within the community. As noted, the effectiveness of this application has been demonstrated in the past literature, however, most investigations utilised either cleaned and sorted post-consumer waste glass or post-industrial waste glass. Our intention was to develop a methodology and product that could be easily replicated in Chembe, with minimal resources. Nonetheless, our results were consistent with those found in the literature, with the waste glass additions contributing to the compressive strength of the brick. Specifically, our results demonstrated an increase in strength compared to the control sample, with the optimum substitution percentage and particle size being 5% of small particle size WG. As a result, the glass addition can be considered as a performance enhancer. Furthermore, although the average water absorption percentage of the burnt clay bricks showed minor fluctuations compared to the control sample, all of the samples performed within the acceptable range.

All samples yielded a compressive strength suitable for Chembe’s architecture (greater than 10 MPa), which overwhelmingly consists of single-story buildings. For this reason, a substitution percentage of 10% WG with large particles is recommended, as it exceeds both the control sample and the required strength of the building applications assessed in this study, while maximising glass inputs, and minimising required crushing. As a circularity model for Chembe, we calculate that, at current rates of glass consumption and brick manufacture, it would take 586 days to fully utilise the waste glass accumulated over 20 years within the community, not accounting for future wastage.

Future research could assess other parameters affecting the properties of the material, such as the effect of the calcination temperature during the during process, the strength/material properties of the bricks at higher substitution percentages and finally the properties of burnt clay bricks with the substitution of very fine particles of WCG (particle size less than 4750 microns). Finally, the firing process itself, which relies on wood unsustainably sourced from local forests (a harvest which has received some scholarly attention (cf. (Abbot 1996; Abbot and Mace 1999)), remains a space for innovation.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ).

Notes on contributors

Noredine Mahdjoub

Dr. Noredine Mahdjoub is a Senior Research Scientist for the South-African Research Chair (SARChI) Waste and Climate Change at the University of KwaZulu Natal in Durban, South-Africa.  He has a background in material science and corrosion engineering with current interests in the Circular Economy, Green Economy, Energy and Hybrid Materials, and completed his PhD in Engineering and Material Science at Manchester Metropolitan University in 2010.  He has also worked for 8 years as a Forensic Engineer and Consultant Engineer in material science. Over the past 5 years, Dr Mahdjoub has exclusively specialised in topics such as alternative building materials, circular economy, urban mining and waste management.

Marc Kalina

Dr. Marc Kalina is a Senior Research Scientist for the South-African Research Chair (SARChI) Waste and Climate Change at the University of KwaZulu Natal in Durban, South-Africa. He holds a B.A. in International Relations from Michigan State University (USA) and a M.A. in Development Studies from Chulalongkorn University (Thailand). He completed his PhD in Development Studies at the University of KwaZulu-Natal in 2017. His research interests centre on human-waste relationships, particularly in the Global South, and he specialises in creative and participatory qualitative methodologies.

Alex Augustine

Alex Riel Augustine completed his  BSc. in Civil Engineering at the University of KwaZulu-Natal Howard College in 2020. Currently, he is a practicing Civil and Structural Engineer in KwaZulu-Natal.

Elizabeth Tilley

Prof. Elizabeth Tilley was a Senior Lecturer in the Department of Environmental Health at the University of Malawi, the Polytechnic in Blantyre, Malawi from 2015-2020, and is currently an Associate Professor in the Department of Mechanical and Process Engineering at the ETH Zurich as well as being an Honorary Research Fellow at the SARCHI Chair in Waste and Climate Change at the University of KwaZulu-Natal, Durban, South Africa.  As an engineer and an economist, Elizabeth is interested in the technological, social financial drivers for sustainable urban services in low-income settings.  Her current work is focused on the impacts of solid waste management on sanitation systems, air quality, and wellbeing in the growing cities of the Global South.

Notes

1. Alan, 04/09/2019.

2. Alan, 04/09/2019.

3. The information presented here draws on a number of informal, loosely-structured interviews held with brick-makers and other stakeholders within Chembe in September 2019.

4. Some brick makers described using sand, while others did not.

5. The largest clay bricks manufacturer in South Africa.

6. The value is based on 6 in order to dilute the mass fluctuation.

7. The bricks were tested in a Rohloff Automatic Compression Testing Machine.