Bio-plastics in the context of competing demands on agricultural land in 2050

Abstract

Recent trends in the bio-plastics industry indicate a rapid shift towards the use of bio-derived conventional plastics such as polyethylene (bio-PE). Whereas historically a significant driver for bio-plastics development has been their biodegradability, the adoption of plastics such as bio-PE is driven by the renewability of the raw materials from which they are produced. The production of these renewable resources requires the use of agricultural land, which is limited in its availability. Land is also an essential requirement for food production and is becoming increasingly important for fuel production. The research presented in this paper envisages a situation, in the year 2050, where all plastics and liquid fuels are produced from renewable resources. Through the development of different consumption and productivity scenarios, projected using current and historic data, the feasibility of meeting global demands for food, liquid fuels and plastics is investigated, based on total agricultural land availability. A range of results, comparing low-to-high consumption with low-to-high productivity, are reported. However, it is from the analysis of the mid-point scenario combinations, where consumption and productivity are both moderate, that the most significant conclusions can be drawn. It is clear that while bio-plastics offer attractive opportunities for the use of renewable materials, development activities to 2050 should continue to focus on the search for alternative feed stocks that do not compete with food production, and should prioritise the efficient use of materials through good design and effective end-of-life management.

Keywords:

• bio-plastics
• land use
• biomass materials
• sustainable materials
• managing use and consumption

1. Introduction

Although the first synthetic plastic material was unveiled in 1862, it was the discovery and subsequent commercialisation of polyethylene (PE) in the 1930s that triggered rapid growth in plastics use (American Chemistry Council 2010). In 2008, the global production of plastics was around 245 million tonnes with the most significant end uses being in packaging (38%) and construction (21%). Almost half of the total plastics consumed takes the form of PE and polypropylene (PlasticsEurope 2009). Plastics are typically made from hydrocarbon monomers: products obtained from the cracking of crude oil and natural gas. Estimates state that the production of plastics accounts for around 4%–5% of total crude oil consumption (Queiroz and Collares-Queiroz 2009).

1.1 The role of plastics in a sustainable society

The role of plastics in a sustainable society is often held in question. The non-renewable nature of fossil fuel feed stocks and the persistence of plastics waste in the environment present a negative image in terms of resource consumption and end-of-life management. In addition, the primary application of plastics is in packaging, which as a highly visible and high-volume waste stream has become almost symbolic of our consumer society's perceived excesses and wastefulness. The reality, however, is more complex. Plastics often offer many benefits over alternative materials, with versatility, low weight and high durability being distinctive characteristics. In particular, plastics packaging can help reduce emissions from transportation of food by weight reduction, and offers the potential for substantial reduction in food waste (Advisory Committee on Packaging 2008). The thermoplastic nature of the majority of polymers used in packaging means that recycling can be readily achieved, with 54% of the post-consumer plastics being directed to energy recovery and recycling operations in Europe in 2009 (PlasticsEurope 2010).

1.2 The development of bio-derived plastics

Biopolymers or bio-derived plastics (BDPs) are polymeric materials which, in contrast to conventional plastics, are produced from renewable resources. Some of the first plastics were manufactured from cellulose, but it has only been within recent decades that a real drive to develop new BDPs has emerged.

Initial efforts concentrated on the development of plastics that were both bio-derived and biodegradable. Biodegradable plastics offer potential for alternative end-of-life management processing (Song et al. 2009), including the recovery of soil nutrients through composting or the recovery of nutrients and energy through anaerobic digestion. Perhaps the most commercially advanced biodegradable BDP is polylactic acid (PLA), derived from starch. PLA has similar properties to polyethylene terephthalate (PET; Auras et al. 2006) and finds commercial application in a range of packaging types, including bottles, trays and clamshells (NatureWorks LLC 2011). Other biodegradable BDPs include thermoplastic starch and polyhydroxyalkanoates. While significant interest has been demonstrated for the application of these materials in packaging, BDPs are also suitable for higher-value applications including electrical and electronic equipments and within the automotive industry. Although promising, these materials are still immature in their development, such that their performance and cost have limited commercial uptake (Crank et al. 2005, Shen et al. 2009).

More recently, a growing range of conventional polymers is being produced (in full or in part) from ethylene, derived from bio-ethanol. These polymers include bio-derived polyethylene, bio-derived PET and bio-derived polypropylene. These BDPs are functionally identical to their fossil-derived counterparts, and so are compatible with existing manufacturing and recycling processes. Figure 1 shows the global growth in the capacity for the manufacture of BDPs in recent years, and illustrates a growing trend in the uptake of these non-biodegradable BDPs (Colwill et al. 2009).

Figure 1 Global production capacity for compostable (biodegradable) and non-compostable BDPs (European Bioplastics 2009).

1.3 Demands and constraints on renewable resources

The data presented in Figure 1 illustrate an increasing emphasis on renewability as opposed to biodegradability with regard to the development of BDPs. However, the benefits of renewability are only realised for as long as the supply of renewable resources required for BDP production exceeds demand. Increasingly, emphasis is being placed on the use of crop-based materials as alternatives to fossil fuels across a range of applications, including for the production of bio-ethanol and bio-diesel as liquid fuels for transportation.

Concerns over competing demands on agricultural land have led to various studies on the impacts of bio-fuel production on food supplies (e.g. Escobar et al. 2009, Rathman et al. 2010, Van der Horst and Vermeylen 2010, Ajanovic 2011, Cai et al. 2011, Harvey and Pilgrim 2011). Evidence of localised price increases for agricultural land as a direct result of the introduction of energy crops is cited by 11 authors in a review conducted by Rathman et al. (2010). However, the review reports that a similar number of studies dispel the idea of food and fuel crops being in competition for land resources. The majority of studies in this area are concerned primarily with bio-fuel production, and few consider within their scope the production of additional products (i.e. plastics) from these renewable resources. A common feature of all futuristic studies is the uncertainty that lies within the projections of human consumption patterns and land productivity (Gerbens-Leened and Nonhebel 2002, Wolf et al. 2003).

2. Research aim and methodology

The primary aim of the research presented in this paper is to investigate the availability of land for the production of BDPs in a future scenario where fossil fuel resources have been exhausted. Although it is unrealistic to suggest that this scenario will be fully realised by 2050, it is generally accepted that within this time frame oil and gas resources will become seriously constrained (Shafiee and Topal 2009, WWF 2010). We therefore examine an extreme situation, where all plastics and liquid fuels (petrol and diesel) are produced from agricultural crops. In addition, we assume that the land available must also support food production. Production of fuel for stationary power generation is not considered in our research, based on an assumption that existing technologies, including nuclear and renewable energy, will be available as alternatives to biomass-based technologies.

In order to conduct the research, three consumption scenarios have been developed based on projected requirements for the year 2050. In Section 3, we identify the key parameters used to define these scenarios, which include global population, food requirements, liquid fuel requirements and demand for plastics. Historic trends are used to project consumption patterns to 2050. In addition, parameters affecting productivity, namely land availability and agricultural yields, are identified and evaluated. The data developed in Section 3 are used to define a range of scenarios for consumption and productivity, which are, in turn, used to address the primary research aim. HIGH, MID and LOW consumption scenarios are defined in Section 4, covering a range of possible situations for the year 2050. In addition, HIGH, MID and LOW scenarios are defined for productivity based on a range of possible average crop yields for the year 2050. Section 5 presents the results generated from the analysis of these scenarios. Total land requirements to support the production of food, liquid fuels and plastics are evaluated for each of the productivity scenarios, in combination with the HIGH, MID and LOW consumption projections. These values are compared with total land availability in order to demonstrate the feasibility of substituting the use of fossil fuel resources with renewable crops for these applications. The discussion of the results generated from the analysis includes identification of significant factors that, although outside the scope of the research presented in this paper, will also impact upon land availability and productivity. Finally, some research conclusions are presented in Section 6.

3. Evaluation of key parameters to support scenario definition

Consumption and productivity scenarios, defined in Section 4, have been developed based on historic trends and existing data for global population, human food requirements, demand for liquid fuels and plastics, land availability and agricultural yields. These data were used to generate projections to the year 2050, as described in Sections 3.1–3.5.

3.1 Global population

Global population is one of the main factors that will impact the demand for resources in the future. The projections used to estimate global population in 2050 were based on statistics for the years 2002 and 2008 (Central Intelligence Agency 2002, 2008). Using three alternative growth scenarios, high-, low- and mid-range projections were calculated, and are shown in Figure 2.

Figure 2 Projections for global population growth to 2050. High, mid and low projections used in the research are plotted against a selection of projections reported in the literature (FAO, Food and Agriculture Organisation of the United Nations 2011; USCB, US Census Bureau 2011; UN 93, Haub 1994; UN 98, United Nations 2008; PAI, Young et al. 2009).

The mid-range projection was calculated, based on the percentage growth in population for the period 2002–2008. The global average growth rate was calculated to be 7.14% for this 6-year period, although the growth rate for individual countries varied considerably. In order to calculate our mid-range projection, constant growth rates are assumed to 2050 for all countries having a growth rate equal to or below the global average (7.14%) for the period 2002–2008. For countries whose growth from 2002 to 2008 exceeded the global average, a growth rate of 7.14% is assumed for each subsequent 6-year period to 2050. This results in a 42% increase in global population between 2008 and 2050.

A low-range projection for global population to 2050 was also calculated, based on an extrapolation of the growth rate for the period 2002–2008. In this projection, the basic growth rate for an individual country over a 6-year period is capped at 20%. Using this basic growth rate, additional factors are incorporated for each 6-year period in order to represent a steady decline in growth rate, with the global population peaking around 2030. Following 2030, the global population enters a period of gradual decline. These assumptions are consistent with theories presented in the literature (United Nations 2004). In this projection, the global population in 2050 is estimated to be around 7.5 billion, which represents an increase by 10% from 2008.

Similarly, a high-range projection for global population was calculated. In this projection, countries' individual growth rates for 2002–2008 were assumed to remain constant. Only countries with a growth rate greater than 40% had their projected growth rate reduced by 20% for every 6-year period. These countries were identified in general as being young economies with populations in 2008 of below 5 million people.

While the general consensus of opinion leans towards a gradual slowdown in the rate of global population growth, there are other more polarised views that predict either a population collapse to around 2 billion people (Duncan 2001) or a continued acceleration in growth driven mainly by developing countries, which could see world population reach 13 billion by 2050 (Dahl 2010). The high and low figures used in our calculations are more conservative and in line with more widely accepted worst- and best-case scenarios, as shown in Figure 2.

3.2 Food requirements

In order to calculate the food requirement of the population in 2050, we must consider two key factors: population size (discussed in Section 3.1) and population diet. It is common practice in studies of this nature to express the wide variety of foodstuffs that make up the human diet in a single unit of measure. Considering the diversity of animal- and plant-based materials produced globally and the wide range of farming methods used, there will always be limitations whatever system is employed. For the purposes of this research, it was decided that a physical measure such as kg wheat equivalent (Nonhebel 2005) or grain equivalents (GE; Penning de Vries et al. 1995) would be more appropriate than less tangible values such as calories or joules.

A simple method for estimating the average global diet was followed, based on three diet types described by Penning de Vries et al. (1995) and expressed in GE. These are vegetarian (low GE), moderate (mid GE) and affluent (high GE). The diet types reflect both the amount and type of food consumed. The vegetarian diet describes an ample and healthy diet of grains, tubers, crops and pulses with some milk. The moderate diet includes a small amount of meat and dairy produce similar to that of Japan or Italy, while the affluent diet is found in rich societies, such as the USA, and includes food for pets. Our projections to 2050 assume values for average annual food requirement per capita shown in Table 1. These average values present an image of an equitable society, where food is equally distributed. In reality, it is likely that current inequalities will persist.

Table 1 Definition of LOW, MID and HIGH consumption scenarios, based on projections to the year 2050.

		Consumption scenarios 2050
		LOW	MID	HIGH
Total global population		8 × 10^9	9.5 × 10^9	12 × 10^9
Food requirement (GE per capita, kg)		475	875	1530
Liquid fuel requirement (total, litres)	Petrol	2.6 × 10^12	3.4 × 10^12	4.1 × 10^12
	Diesel	1.9 × 10^12	2.5 × 10^12	3.0 × 10^12
Plastic requirement (total, kg)		3.4 × 10^11	7.0 × 10^11	13.0 × 10^11
Note: Food requirement data taken from Penning de Vries et al. (1995). Calculation of all other data projections is detailed in Section 3.

These values are a simplistic reflection of food consumption based around primary food types: meat, dairy and plant. They do not take into account the resources required for the subsequent distribution and processing of food, wastage and spoilage levels or the production of beverages and luxury goods.

3.3 Demand for liquid fuels

Although coal and gas are mainly used for heat and power generation, the majority of crude oil is used in liquid fuels for transportation (Energy Information Administration 2011). In our research, it was assumed that existing alternative technologies, such as nuclear, solar, wind or wave power, could be used to generate sufficient stationary power to meet human demand in 2050. Transport fuels such as diesel and petrol are highly concentrated forms of relatively safe portable energy for which a large infrastructure and support system exists. Here bio-fuels offer a viable alternative at present as they allow for continued use of existing products and infrastructure, in contrast to, for example, electric cars (Dufey 2006). Ethanol is already added to petrol in many countries at levels of around 2%–3%; however, national targets seek to increase this to as much as 10% by 2020 (Dufey 2006). Fuel blends containing up to 85% ethanol are currently available for use in specially designed vehicles (Corts 2010). The two liquid fuel groups, diesel and petrol, are considered separately, as different crops are used in their manufacture. Bio-diesel is produced from oil crops and bio-ethanol from sugar/starch crops. In the scenario analysis for 2050, it is assumed that bio-ethanol replaces petrol and bio-diesel replaces diesel. Figure 3 shows current and future demand trends for these two fuel groups. Three consumption projections for 2050 were used to give a high, mid and low figure. These were calculated based on data from the Organisation of the Petroleum Exporting Countries (OPEC 2009) for projected oil consumption to 2030. For the low projection, continued growth in line with OPEC estimates is assumed until 2020, at which point further growth ceases. The mid-range projection follows OPEC estimates to 2030 and extrapolates this growth rate to 2050. These OPEC estimates reflect the slow down in growth which has occurred since 2008. The high projection uses historic data (Energy Information Administration 2009) to calculate the higher growth rate experienced prior to the recession in late 2008. This higher rate of growth was applied from 2010 to 2050 based on the assumption that growth returns to pre-2008 levels and that supply will keep pace with increased demand.

Figure 3 Global demand for liquid fuels, projected to 2050. Petrol and diesel account for around 75% of global crude oil demand. Original data sources and projection calculations are detailed in Section 3.

3.4 Demand for plastics

The demand for plastics has increased annually since the 1950s. Three consumption estimates for 2050, high, mid and low, were calculated (Figure 4) based on historic data for world production of plastics from 1950 to 2005 (PlasticsEurope 2009). The low projection assumes continued growth at a rate of 4.3% for every 5-year period, in line with the level of growth observed between 2005 and 2009. This low growth rate follows the fall in demand during the recession of 2008 and 2009, offset in part by the rise in demand during the rest of this 5-year period. The mid-range projection assumes a growth rate of 14% every 5 years, based on the average growth rate observed between 2000 and 2010. The high-range projection uses a 5-year growth rate of 23%, which was calculated as the average growth observed for each 5-year period from 1990 to 2010.

Figure 4 Global demand for plastics, projected to 2050. Original data sources and projection calculations are detailed in Section 3.

Our projections for plastics consumption to 2050 are inclusive of all plastics currently in use. The two main families of plastics are termed thermoplastics and thermosets, of which thermoplastics accounts for the largest share. The substitution of the current range and diversity of polymers in use with an equivalent BDP is a complex scenario. The research simply assumes that a range of BDPs will be available to meet the technical requirements in 2050. Bio-PE was identified as a representative BDP on which to base calculations for land requirements to support plastics production. PE in its various forms: high density, low density and linear low density, is currently the largest and most widely used polymer. Given the trends identified in Figure 1, it also seemed reasonable to select bio-PE as a reference. In the discussion of yields (Section 3.6), we describe the land requirements for bio-PE in the context of other BDPs, and further justify this approach.

3.5 Land availability

The production of food is the largest industrial use of both land and water (Wallace 2000, Gerbens-Leened and Nonhebel 2002, Naylor et al. 2005), yet the land available that is suitable for food production is limited. Of the 30% of the earth that is not under water, only around 31% is suitable for arable crops and 33% for grazing (Penning de Vries et al. 1995). Other estimates suggest that less than half of the world's land area (3000 million ha) is suitable for agricultural use, which includes grazing, with the majority of this productive land already in use. Further expansion would be limited at the most to around 500 million ha and this would be achievable only through deforestation (Kindall and Pimentel 1994).

For the purposes of this research, land availability data were based on statistics available from the United Nations (Food and Agricultural Organisation of the United Nations 2011). Three classes of land were identified as being potentially available for growing crops, suitable for food, bio-fuel and/or BDP production. These were ‘crop land’ (including all arable land and permanent crops), ‘grazing land’ (including all permanent meadows and pastures) and ‘forest land’. By plotting global land use statistics from 1950 to 2010, it was observed that in comparison with population growth during the same period, the increase in cultivated land use through gradual deforestation has been modest (Figure 5). It was therefore decided that current land use data would be used to reflect land availability in 2050.

Figure 5 Historic data for land use (Food and Agricultural Organisation of the United Nations 2011) in comparison with global population growth, between 1950 and 2010.

3.6 Agricultural yields

The demands on our planet's resources from its human inhabitants have already exceeded the Earth's bio-capacity by approximately 50%. This overshoot, however, is largely attributed to the rise in CO₂ emissions, which have grown by 20-fold since 1961, and currently account for more than half of this global ecological footprint calculation (WWF 2010). These CO₂ emissions are primarily the result of the rapid increase in the use of fossil fuels, particularly crude oil, during the latter half of the twentieth century (Ewing et al. 2010). The significance of the increased use of fossil fuels to agricultural yields can be realised when one considers that since the 1950s the area of land use for agriculture, such as the growing of cereal crops, has remained relatively constant, while the human population has more than doubled (Figure 5). Although a number of factors have contributed to the success in raising agricultural yields, the increased use of fossil fuels has been significant in making current intensive farming practises possible. As land is ultimately a finite resource, improving yields is the most obvious means of meeting increased demand.

Yields can vary significantly depending on the quality of the land, type of farming practice, water availability, additional fertilizer used, climate and type of crops grown, etc. In some areas (e.g. the tropics) up to three harvests per year can be achieved. Using a standard measure of GE, yields can vary from under 1 tonne per ha per year in developing countries to over 9 tonnes per ha in the USA and Brazil. In 2010, the global average was around 4.6 tonnes per ha per year. Although a single GE figure can provide a useful standard for making comparisons between global consumption and production levels, it can be misleading when comparing different land and crop types. To avoid over-simplification, high-, mid- and low-yield scenarios for each of the key resource groups have been developed and comprise food, liquid fuels and plastics. The base data used for these yield scenarios were tailored to each resource group and reflect the crop and land types that would be used.

3.6.1 Food yields

For food, actual yield statistics for cereal production in 2009 were used (Food and Agricultural Organisation of the United Nations 2011). The mid-yield figure took the global average for this year; the high-yield value took the average for the USA and the low-yield value took the average for India. Achieving average USA yields at the global level might appear to be an overly optimistic projection for 2050, even for the high-yield value. However, when considering the historic trend in increased yields over the past 50 years (Figure 5), it may not be unreasonable to use this projection. The low-yield figure used India as representing a range of agriculture systems, land types, crops and climates. It is not excessively low and reasonable as a low global figure when considering the potential impact of using less productive land, water shortages, fertiliser and fuel limitations and the possible effects of climate change.

3.6.2 Liquid fuel yields

Liquid fuels calculated bio-diesel and bio-ethanol separately due to the variation in yields achieved from the different types of crops used in their manufacture. The mid, low and high values are based on actual 2009 average yields achieved in litres per m² for ethanol and bio-diesel (Sanderson 2006, Singh et al. 2011).

For bio-diesel, the low-yield figure is based on average yields from rapeseed crops. The high-yield figure is based on production of bio-diesel from Jatropha. The mid-yield value was calculated as the average of these two extremes.

For bio-ethanol, the low-yield value is based on corn as the feedstock using the lower end of the data range reported in the literature. For the high-yield value, data representatives of bio-ethanol produced from sugar cane and switch grass are used, taking the average of the higher values reported. The mid-yield value is taken as the mid-point between the high- and low-yield values and compares closely with the average yields obtained from switch grass, the high end of corn and the low end of sugarcane.

3.6.3 Plastics yields

The low-, mid- and high-yield values for the production of BDPs are based on current production data for bio-PE from ethanol. Low-, mid- and high-yield values for bio-ethanol production (Section 3.6.2) were combined with a PE yield of 1 kg from 2.3 L of ethanol (Braskem 2010). This provided a yield, expressed in terms of kg BDP produced, per m² of land.

In terms of the production of BDPs in general, bio-PE was identified as being relatively resource inefficient. For comparison, current production figures indicate that 4 kg of wheat starch will produce approximately 2.9 kg of PLA but only 1.1 kg of PE (Siebourg and Schanssema 2008). Given that it is not possible to accurately predict which BDPs and what percentages of each will contribute to the total plastics demand in 2050, it was decided that to select the more resource-demanding PE would provide a ‘worst-case’ view of land requirements. This decision was also underpinned by the data shown in Figure 1, which indicates the relative growth of non-degradable BDPs compared with biodegradable BDPs. In terms of material substitution, PE is the dominant polymer type currently in use and it is known that bio-PE can substitute conventional PE without any loss in performance during processing, use and at end-of-life.

4. Scenario definition

Based on the projected data described in Section 3, a range of scenarios have been developed in order to explore future land availability for the production of plastics in a society reliant on renewable resources.

Three consumption scenarios are defined in Table 1. The parameters defined for each scenario are global population, food requirements and demand for liquid fuel and plastics. Food requirement is defined per capita, whereas projections for liquid fuel and plastics are based on data for total global demand. All data are defined for the year 2050. The three consumption scenarios defined in the research are as follows:

•

LOW consumption In the LOW consumption scenario, global population growth peaks at 2030 and then declines slowly to 2050. The average diet is low in animal produce and high in grain. Total global demand for liquid fuel has remained at present-day levels, reflecting increasingly prohibitive costs associated with motoring and increasing availability of alternative and more efficient transportation technologies. Demand for plastic has shown only marginal growth, as a result of poor economic growth and/or improved material efficiencies through good design and effective use of recycling.

•

MID consumption In the MID consumption scenario, the global population continues to grow at current rates to 2050. Average eating habits include more animal produce than in the LOW consumption scenario, reflecting economic growth in the developing world. Demand for liquid fuel has also continued to grow at current rates, with increased demand from the developing world counterbalanced with improved efficiencies and the adoption of alternative technologies in transportation by developed countries. Growth in plastics usage has also been moderate.

•

HIGH consumption In the HIGH consumption scenario, the rate of population growth to 2050 has been increasing more dramatically than in the MID consumption scenario. Economic growth in developing countries is reflected in a spread of consumerism and the adoption of western lifestyles. This has resulted in an increased level of animal produce in the average diet, increased demand for liquid fuel and escalated demand for plastics. Sustainability concerns have had little impact on consumption patterns. Whereas consumption scenarios are used to identify potential demands on land in 2050, the availability of renewable resources is defined by productivity scenarios. Based on the data explored in Section 3, the amount of land available is assumed to remain constant for the LOW, MID and HIGH productivity scenarios. Average agricultural yield varies for each scenario, as described below:

•	LOW productivity The LOW productivity scenario in 2050 is defined by poor yields, which are lower than the average global yields achieved today. This scenario could arise as a result of exhaustion of previously productive agricultural land and reduced availability of fertilisers. Intensive farming practices have been slow to spread to the developing world and unpredictable weather patterns have had localised catastrophic impacts on crops.
•	MID productivity The MID productivity scenario in 2050 is defined by moderate yields achieved through a maintenance of current farming standards. Increased yields from the spread of intensive farming practices are counter-balanced by the exhaustion of land in over-cultivated areas.
•	HIGH productivity The HIGH productivity scenario in 2050 is defined by high yields, above current average values, achieved through a mixture of good land management, effective crop selection and improvements in agricultural practice. Developing countries adopt more intensive farming practices, with increased use of fertilisers and mechanised processes.

5. Scenario analysis and discussion

The scenarios developed in Section 4 have been used to investigate the feasibility of meeting the global demand for plastics entirely from the use of agricultural crops, thus competing with the production of food and liquid fuel. Sections 5.1–5.3 present the results generated based on the HIGH, MID and LOW productivity scenarios defined in Table 2. In each section, calculated total land requirements to support the LOW, MID and HIGH consumption scenarios are presented and compared with total land availability. Section 5.4 presents a discussion of the validity of the results generated, by identifying some limitations to the current research.

Table 2 Definition of HIGH, MID and LOW productivity scenarios, based on projections to the year 2050.

	Productivity scenarios 2050
	HIGH	MID	LOW
Average agricultural yield (kg GE m^− 2)	0.72	0.35	0.25
Average bio-ethanol yield (L m^− 2)	0.80	0.55	0.30
Average bio-diesel yield (L m^− 2)	0.20	0.15	0.10
Average BDP yield (kg m^− 2)	0.35	0.24	0.13
Total land availability (m^2)	8.7 × 10^13
Crop land (m^2)	1.6 × 10^13
Grazing land (m^2)	3.4 × 10^13
Forest (m^2)	3.8 × 10^13
Note: Land availability is assumed to be constant for all scenarios. Calculation of data projections is detailed in Section 3.

5.1 HIGH productivity scenario analysis

The total land requirement to support human demand for food, liquid fuels and plastics was calculated for each consumption scenario defined in Table 1, using the HIGH productivity scenario defined in Table 2. The assumption is that the total demand for petrol and diesel fuels is met by bio-ethanol and bio-diesel, respectively, and the total demand for plastics is met by BDPs. The results from these calculations are shown in Figure 6. Total land availability is shown for comparison.

Figure 6 Scenario results for LOW, MID and HIGH consumption scenarios in combination with HIGH productivity.

This set of results indicates that in a HIGH productivity scenario, it is feasible that human demands for liquid fuels and plastics could be met by using renewable raw materials, without significant threat to food production. Even for the HIGH consumption scenario, the majority of food requirements could be met by using crop land, with some food requirements being met by the use of grazing land for the production of meat and dairy. A portion of crop land would therefore remain available for the production of liquid fuels and plastics, with the remaining demand for liquid fuels and plastics being met by grassy crops grown on grazing land. The total land requirement for plastics production is between 5% and 7.5% of the total land required to support these competing end uses.

The combination of low consumption and high productivity shown in Figure 6 is indicative of the ‘best case’ scenario developed in the research. This scenario assumes low global population and a radical shift in average human behaviour towards a diet, which is low in animal produce, and demand for liquid fuel and material similar to current consumption rates. In addition, the yield assumed for the HIGH productivity scenario is in line with current yields in the most advanced farming communities.

5.2 MID productivity scenario analysis

Figure 7 shows the total land requirements for LOW, MID and HIGH consumption scenarios in combination with the moderate yields defined in the MID productivity scenario. It can be seen from the results that even for the LOW consumption scenario, demand for land exceeds the available crop land and utilises almost half of the available grazing land. The total land requirement for the MID consumption scenario is similar to the total land requirement for the HIGH consumption scenario in combination with HIGH productivity (Figure 6). For the MID consumption scenario, the land requirement for food, liquid fuels and plastics totals all available crop and grazing land. For the HIGH consumption scenario, the total land requirement extends to an area as large as the entire crop and grazing land, as well as the majority of the forest land.

Figure 7 Scenario results for LOW, MID and HIGH consumption scenarios in combination with MID productivity.

This MID productivity scenario reflects average crop yields achieved today, and as such presents a scenario that could be realistically envisaged. It is likely that some improvements will be made in crop yields in the developing world, and these would counterbalance the reductions in crop yields elsewhere in the world through soil degradation and land exhaustion. The results for the MID consumption scenario presented in Figure 7 reflect the mid-point developed in this research, which is possibly the most realistic or likely situation for 2050. The results here suggest that, on the basis of the assumptions adopted in the calculation of land availability, a switch to crops as raw materials for liquid fuel and plastics cannot be dismissed as being totally unfeasible. The total land requirement falls marginally within the total area of crop and grazing land available. This result highlights the importance of effective resource management, in both agricultural production and in consumer behaviour. The results for the HIGH consumption scenario here illustrate the impact of uncontrolled growth in demand for fuel and materials and the effect this would have on the ability to meet demands by the use of renewable resources. It is unfeasible to suggest that the complete destruction of forest land to support food, fuel and plastics production provides a sustainable solution to meeting human needs. In addition to playing an important role in supporting the planet's ecosystems, forests provide an essential source of wood and charcoal fuels, as well as raw materials for other industrial uses. The results presented in Figure 7 emphasise the importance of decoupling economic growth with increasing consumption: the principal challenge of sustainable development.

5.3 LOW productivity scenario analysis

Figure 8 shows the LOW productivity scenario and the resulting land requirements for LOW, MID and HIGH consumption scenarios. Low crop yields cause demand for land to significantly exceed available crop land for all three consumption scenarios. For the MID consumption scenario, a large proportion of forest land would be required to meet the human demands considered within the research, and for the HIGH consumption scenario, land requirements could not be met, even supposing all forest land could be cleared and used for agricultural purposes.

Figure 8 Scenario results for LOW, MID and HIGH consumption scenarios in combination with LOW productivity.

The results presented in Figure 8 for the HIGH consumption scenario illustrate the ‘worst case’ developed in this research, in which land availability is not sufficient to meet food requirements, and therefore provides no opportunity for providing crop-type resources for competing markets. As with the ‘best case’ presented in Section 5.1, the likelihood of this ‘worst-case’ scenario being realised is low. The low crop yield defined in the low productivity scenario used as the basis of these calculations could only be envisaged as a result of the extreme effects of climate change or some other catastrophic occurrence. However, this extreme scenario presents a picture of a situation where consumption patterns remain unchecked and a lack of concern for the environmental impact of human behaviour results in substantial degradation of the planet's resources.

5.4 Limitations of the scenario analysis

The scenarios developed in this research, and the results presented in Figures 6-8, are intended to provide a broad view of the situation regarding the availability of land in terms of providing renewable resources as raw materials for liquid fuel and plastics. The variation in the results presented, from the ‘best-case’ to ‘worst-case’ scenarios, indicates the complexity of the issue, as well as the sensitivity of the situation to factors such as population growth and crop yields, which are difficult to predict. Some of the issues that have not been directly included within the research, but are acknowledged as being significant, are identified below.

In defining the consumption scenarios, it has been assumed that the only demands on agricultural land will be food, liquid fuels and plastics. Other significant uses include the growth of tobacco crops and the production of natural fibres, such as cotton, for textiles. Some industrial processes, such as steel production, consume substantial quantities of coal, which in future may need to be substituted. The production of stationary power (e.g. in power stations) has been deliberately excluded from the scope of the research, while in reality there a likelihood that some stationary power will be generated using biomass grown specifically for that purpose. As the global population grows, it may also be that some agricultural land area is lost to the construction of roads and homes. Furthermore, the use of forest land for solid fuel production (wood and charcoal) and other industrial purposes has not been incorporated in our considerations with respect to future projections.

We have also based our projected consumption requirements on historic and current human behaviours. In reality, it is understood that human behaviour changes over time and adjusts, in particular, to economic and social factors. Although the consumption scenarios developed in the research encompass a range of potential situations for the year 2050, we are not able to predict step changes in human behaviour that could radically change the demand for liquid fuels and/or plastics.

In defining the productivity scenarios, we have taken a rather simplistic approach in developing average crop yields based on data reported in the literature. In reality, agriculture is heavily dependent on a complex list of factors, including water availability, climate, weather patterns and the availability of fertilizers, machinery and other infrastructure required to support farming. In particular, the availability of clean drinking water is essential for human survival, and the redistribution of water for irrigation can have catastrophic impacts on local communities. In our research, we have made the assumption that sufficient water is available to agricultural land. This assumption is unlikely to reflect the real situation in 2050. The nature of agriculture is such that the production of renewable resources is closely linked with the weather and the climate. Global changes in climate have the potential to substantially change agricultural yields, as well as presenting the possibility of rising sea levels and the consequent loss of low-lying arable land. Extreme weather conditions, such as droughts, hurricanes and floods, can have catastrophic impacts on farming and these perhaps take on even greater significance as land availability is stretched. Even without such extreme events, the production of raw materials from agriculture, where availability is so closely linked with the seasons and fluctuations in weather, is characteristically different from the relatively constant business of extracting fossil fuels. The resulting impacts on trade and economic behaviour have not been considered in this research.

On a more positive note, it is possible that alternative sources of raw materials may be developed to support the production of liquid fuels and plastics. Already, a shift towards the use of cellulosic materials, rather than sugars and starches, is planned for both product types. Research into the use of algae to produce biomass is promising, and although farming this resource from the sea may introduce its own environmental problems, there is potential to reduce the strain on land and remove competition for food production. Similarly, opportunities to utilise the resources available from waste have the potential to alleviate the requirement of growing ‘virgin’ crops as raw materials for fuels and/or plastics production.

Finally, we have conducted a theoretical analysis in which global demand has been compared against global supply. In reality, perhaps the biggest challenge associated with food production is not the growth of sufficient crops, but rather the distribution of food to the people who need it. Today, despite there being more than adequate resources available at the global level, it is estimated that over 1 billion individuals live in poverty and hunger (Food and Agricultural Organisation of the United Nations 2009). Simply demonstrating a theoretical ability to meet global demand by no means indicates that the requirements of the individual will be met. The challenge of distribution relates not only to food but also to renewable materials required for the production of liquid fuels and plastics. Transportation of these raw materials from agricultural areas to processing plants to the consumer introduces additional environmental impact and resource demands within the supply chain.

6. Conclusions

The production of plastics from renewable resources at present offers an attractive opportunity for reducing fossil fuel consumption and improving the apparent sustainability of products and packaging. However, in the future, increasing pressure on land for the production of food and liquid fuels will challenge priorities in terms of the allocation of renewable resources. The wide range of scenarios presented in this study illustrates the complexity of the issues involved in predicting human consumption patterns and land productivity in the future. In the worst case (low productivity combined with high consumption), the ability of agricultural land to support human demands is far exceeded, even with the expansion of farming into existing forests. In the best case (high productivity combined with low consumption), human demands could, theoretically, be met with ease. However, these extreme cases represent possible, but unlikely, situations for the future.

The moderate case (mid productivity combined with mid consumption) represents the most likely situation for 2050, and it is from this that the most significant conclusions from the study can be drawn. Here, the maximum available crop and grazing land is used in its entirety to support production of food, liquid fuels and plastics. In reality, considering the simplified approach adopted in the scenario development applied in this study, as well as the unavoidable inefficiencies in agricultural, manufacturing and distribution processes, this moderate case does not represent a sustainable solution.

This failure leads us to conclude that although renewable fuels and materials appear attractive today, they do not provide a straightforward global solution that will allow human consumption patterns to remain unchecked. Although both plastics and liquid fuels are essential requirements of modern supply chains, and will remain so, especially within the context of increased urbanisation and population growth, food production will always remain a priority. This conclusion, developed from an evaluation of global resources and requirements, does not reflect regional variations in local land availability. Regions rich in agricultural land may well be able to support the demands of their local populations into the future. However, as global resources become increasingly constrained, it is debatable whether the priorities of individual countries can remain detached from global pressures.

In terms of the BDP industry, continued emphasis should be placed on the exploration and development of alternative feed stocks for plastics, which do not compete with food production, for example, algae and waste. In addition, improvements in resource efficiency, achieved through the development of efficient recycling processes, innovative design and changed consumer behaviour, will continue to be essential for sustainable development.

Acknowledgements

The research was funded by the EPSRC. The authors would also like to acknowledge helpful comments from external reviewers.