An environmental impact calculator for greenhouse production systems

Journal of Environmental Management 118 (2013) 186e195

Contents lists available at SciVerse ScienceDirect

Journal of Environmental Management journal homepage: www.elsevier.com/locate/jenvman

An environmental impact calculator for greenhouse production systems Marta Torrellas*, Assumpció Antón, Juan Ignacio Montero IRTA (Institute of Agriculture and Food Research and Technology), Ctra. de Cabrils, km 2, 08348 Cabrils, Barcelona, Spain

a r t i c l e i n f o

a b s t r a c t

Article history: Received 8 August 2012 Received in revised form 8 January 2013 Accepted 11 January 2013 Available online 24 February 2013

Multiple web-based calculators have come on the market as tools to support sustainable decision making, but few are available to agriculture. Life cycle assessment (LCA) has proved to be an objective, transparent tool for calculating environmental impacts throughout the life cycle of products and services, but can often be too complex for non-specialists. The objective of this study was therefore to develop an environmental support tool to determine the environmental impacts of protected crops. An effort was made to provide an easy-to-use tool in order to reach a wide audience and help horticulture stakeholders choose efficient options to mitigate the environmental impacts of protected crops. Users can estimate the environmental performance of their crops by entering a limited amount of data and following a few easy steps. A questionnaire must be answered with data on the crop, greenhouse dimensions, substrate, waste management, and the consumption of water, energy, fertilisers and pesticides. The calculator was designed as a simplified LCA, based on two scenarios analysed in detail in previous tasks of the EUPHOROS project and used as reference systems in this study. Two spreadsheets were provided based on these reference scenarios: one for a tomato crop in a multi-tunnel greenhouse under Southern European climate conditions and the other for a tomato crop in a Venlo glass greenhouse under Central European climate conditions. The selected functional unit was one tonne of tomatoes. Default data were given for each reference system for users who did not have complete specific data and to provide results for comparison with users’ own results. The results were presented for water use as an inventory indicator and for the impact categories abiotic depletion, acidification, eutrophication, global warming, photochemical oxidation and cumulative energy demand. In the multi-tunnel greenhouse, the main contributors based on the default data were the structure, fertilisers and auxiliary equipment, whereas, for the Venlo glass greenhouse, the main contributors were energy consumption for heating and, to a lesser extent, the structure. The results were evaluated for alternative options of electricity, fertilisers, pesticides and means of transport, as these areas were found to have potential variability, depending on the characterisation model and datasets used. The resulting calculator is a useful tool to simulate the environmental performance of protected horticultural production systems and is also helpful to growers and advisers for evaluating the efficiency of input reduction options. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Simulation Life cycle assessment Environmental performance Horticultural production system Greenhouse

1. Introduction Life cycle assessment (LCA) is a useful management tool for calculating the environmental behaviour of products, services and activities throughout their life cycle. An LCA provides quantitative and good-quality environmental information with multiple applications in management and in relation to other environmental management tools and concepts (Finnveden et al., 2009). * Corresponding author. Tel.: þ34 902 789 449; fax: þ34 975 533 954. E-mail addresses: [email protected], [email protected] (M. Torrellas), [email protected] (A. Antón), [email protected] (J.I. Montero). 0301-4797/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jenvman.2013.01.011

Producing a full LCA is a complex task due to factors such as the methodological requirements and the huge amount of data required (Mourad et al., 2007). This complexity explains why most LCA studies are conducted by researchers, consultants and other professional LCA experts. Communicating the calculation and interpretation of results to non-specialists is an added difficulty of the LCA. Societal concern for environmental issues has increased the demand for environmental information (Halberg, 2004a). To respond to this concern, multiple web-based environmental-impact calculators have been developed as simplified life cycle management tools to support decision making to find more environmentally sustainable solutions. Some of these calculators are free and are

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195

usually oriented towards a specific industry, such as ecodesign (idc LCA calculator; Okala Life Cycle Analysis calculator), construction (Athena EcoCalculator; Kellenberger, 2010), energy (LiMaS WebSuite; Robust DSC LCA calculator) and waste management (Farreny and Gasol, 2012). In the food industry, different calculators are available for the environmental evaluation of personal consumption habits (Epp and Reichenbach, 1999), sustainable shopping (Pascualino et al., 2009) and industrial activities (Azapagic et al., 2010). The Sustainable Development Online website (http://sdo. ew.eea.europa.eu/) provides access to a large number of software tools, as well as links to other sustainable development websites. Apart from giving information about the environmental impacts of products and activities, these LCA tools aim to provide advice on how to reduce these impacts. Most calculators offer indicators for global warming potential (carbon footprint in kg CO2 eq), but few give results for other impact categories. There are few LCA simple software tools focused on the agriculture sector. Agriculture is a topic of major concern and is moving towards more sustainable practices to reduce associated environmental impacts (Haas et al., 2001; Stoate et al., 2009). Some major sources of environmental damage in agriculture are land use, water use, fertiliser and pesticide application, and energy consumption for heating and machinery. Emission calculators focussing on agriculture can be useful for determining major burdens and implementing corrective action to reduce them. Recent contributions in this field are the Cool Farm Tool (Hillier et al., 2011), a greenhouse gas calculator for crop and livestock production at farm level that takes into account the amount of change in soil C; and Musa software (Amores Barrero et al., 2012) for water assessment of different agricultural production systems. In the particular case of protected horticultural production, resource and waste reduction could be studied to achieve more sustainable production (Pires et al., 2011). In order to address these issues, the main goal of this project was to develop an easy-to-use environmental tool to calculate the efficiency of the inputs used to produce protected crops, based on LCA methodology. This support tool is designed for people working in the horticulture industry such as growers, researchers, technical advisers and authorities. The aim was to provide a tool to improve environmental appraisal of horticultural production out of the scientific community (Halberg, 2004b). The calculator can be used to simulate horticultural production systems to improve their management by helping in the evaluation of different input options and choosing the best one. The calculator is based on a simplified LCA and provides results for several impact categories of interest in agriculture. It is a free-access tool and was developed within the context of the Efficient Use of Inputs in Protected Horticulture Research Project (EUPHOROS, 2008e2012). In this paper, the authors present the environmental calculator and discuss its advantages and limitations. 2. Material and methods The design of this environmental calculator was based on LCA methodology. As the goal of the study was to develop an easy-touse tool, the LCA was simplified following the principle of parsimony, which states that modelling should be done by starting with simple assumptions and only adding complications as they become necessary (Pidd, 1996). The present study focuses on two scenarios representative of the current situation of horticultural production systems in Europe under different climate conditions: (1) a tomato crop in a plastic greenhouse under Southern European climate conditions; and (2) a tomato crop in a glass greenhouse under Central European climate conditions. A full LCA for these two crops was conducted and described in detail in previous analyses (Montero et al., 2011). The

187

initial environmental evaluations showed several environmental problems for each scenario and their data were used as the starting point for the development of this environmental calculator. In this study, we used these two tomato crops as reference systems to model the production systems and calculate and compare results using the calculator. As a result, two spreadsheets were developed for the simulation of horticultural production of protected crops in two kinds of greenhouse structures: (1) a multi-tunnel greenhouse and (2) a Venlo glass greenhouse. As the aim was to make the environmental calculator available to the general public, a web platform was used to increase dissemination and accessibility. The calculator is an EXCELÔ spreadsheet consisting of eight worksheets. Four of these sheets are disclosed to users: Instructions, Input Data, Total Results and Detailed Results. The other sheets are not accessible to users and include the database and the calculation of results. The Instructions sheet includes an introduction to the environmental calculator, guidelines on how to use it and a description of the production systems. Users only have to follow three easy steps to conduct an environmental simulation: eSelect a production system eEnter data on the Input Data sheet eCheck the results on the Total Results and Detailed Results sheets The Input data sheet is a questionnaire. Users are asked to fill in cells with their own data to model the production system under study. The sheet provides default values for reference production systems (Montero et al., 2011) that users can use if they do not have a specific value for their own situation. Data are entered in the specific unit indicated in the Units cell. To increase understanding, the questionnaire is structured in several topics: crop, greenhouse structure, climate control system, auxiliary equipment, electricity consumption, material transport, watering, fertilisers, pesticides and waste management. Comments and drop-down lists are included in several cells to facilitate data entry. In addition, a fertiliser calculator was developed by Dr. Luca Incrocci (Dipartimento di Biologia delle Piante Agrarie, University of Pisa) for the calculation of the total amount of nutrients applied to the crop. The results are presented on two separate sheets to facilitate comprehension, Total Results and Detailed Results. The Total Results sheet shows the total results of the production system by functional unit and in several indicators and impact categories of interest in energy and agriculture processes. The total results for the reference production system are included, as calculated from default values, and can be compared with the results calculated from the data entered by the user on the Input Data sheet. A brief definition of each impact category is included so users can acquire knowledge about the damage that can be caused to the environment. The Detailed Results sheet shows the contribution of the production system to the impact categories, broken up in stages. The user’s results can be compared to the reference situation in the tables and figures. We used an attributional LCA for the environmental assessment that is purely descriptive in accordance with the ISO 14040 standard (ISO-14040, 2006). A mass functional unit (FU) was defined to represent the final function of the production system, i.e. 1 tonne of tomatoes. In the reference system, the FU referred to classic loose tomatoes. The system boundary defined was from raw material extraction to the farm gate, including material waste disposal. Subsequent stages and sales processes were not taken into account, as the aim was to focus the calculator on the improvement of production management. The LCA was simplified by including the most relevant processes and flows in the production system.

188

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195

Following the cut-off criteria of the International Reference Life Cycle Data System Handbook (ILCD, 2010), processes with environmental impact percentages below 5% were omitted when they were not considered relevant for an agricultural production system. In the design of the calculator, an effort was made to simplify the questionnaire by asking users for the minimum data necessary to calculate inventory inputs and outputs. Users enter primary data related to greenhouse dimensions and agricultural operations, such as crop period, crop density and volume of substrate per bag; as well as water, fertiliser, pesticide, electricity and fuel consumption. Default data for the multi-tunnel greenhouse were obtained from reference production systems at the Fundación Cajamar Experimental Station (Fundación_Cajamar, 2008), and default data for the Venlo glass greenhouse were obtained from Quantitative Information for Greenhouse Horticulture (Vermeulen, 2008). Default data on fertilisers in the multi-tunnel spreadsheet were taken from the best available techniques as defined by Muñoz et al. (2008). The growing period was used to calculate the amount of materials in the structure and the amount of substrate. In the multi-tunnel structure, a 52week greenhouse occupation period was used as the default value. Although the tomato crop lasted for 9 months, a 3-month resting period was added for maintenance of greenhouse operations and preparation for the following crop. Secondary data for environmental analysis (Table S1 in the supplementary data) were obtained from the Ecoinvent database 2.2 (Ecoinvent, 2010) for processes such as the manufacture of greenhouse components (e.g. steel, aluminium, glass, LDPE), the production of the substrate and pesticides, the electricity production mix, means of transport and disposal. To simplify the environmental calculator, only the electricity production mix in Europe was used. The LCAFoods database (Nielsen et al., 2003) was used for the production of generic fertilisers N, P2O5 and K2O. The indicators and impact categories selected for environmental assessment include five midpoint impact categories defined by the CML2001 method v.2.04 (Guinée et al., 2002), i.e. abiotic depletion (kg Sb eq), acidification (kg SO2 eq), eutrophication (kg PO3 4 eq), global warming (kg CO2 eq) and photochemical oxidation (kg C2H4 eq); one energy flow indicator (cumulative energy demand, MJ); and one inventory flow indicator (water use, m3). Abiotic depletion is related to the extraction of minerals and fossil fuels that form part of the inputs of the system. Ammonia and nitrate emissions from Nfertilisers make major contributions to acidification and eutrophication, respectively. Inputs such as the metal and plastic in the structure, fertilisers, the substrate and the system’s electricity consumption are major contributors to global warming because of the greenhouse gases released to produce them. Gases contributing to photochemical oxidation formation, e.g. ozone, are very harmful for humans, ecosystems and crops. Cumulative energy demand aims to analyse energy use throughout the life cycle of a product, including direct and indirect uses. The production system is structured in six stages to facilitate the assessment and comprehension of results: greenhouse structure, auxiliary equipment, climate control system, fertilisers, pesticides and waste management (Fig. 1). Each stage takes into account the processes and flows included in the manufacture of greenhouse components, the transport of materials, material disposal and greenhouse management (heating and water, and the consumption of fertilisers, pesticides and electricity). Based on the user’s data and reference production systems, the environmental simulation tool was designed to calculate inputs and outputs for the processes summarized in Table 1 and described below (Torrellas et al., 2012a). 2.1. Structure The multi-tunnel structure is a commercial arched-roof, steelframed greenhouse with a plastic covering and the Venlo glass

greenhouse is made of steel, aluminium and glass. The metal elements include posts, reinforcements, gutters, profiles and ventilators. Recycled metal is used to produce the metal elements in the two scenarios. Plastic insect-proof screens in the multi-tunnel greenhouse and energy screens in the Venlo glass greenhouse are also included in this stage. The amount of materials in the structure is calculated using formulae developed from reference production system inventories (Antón, unpublished results) and the greenhouse dimension data entered by the user on the Input Data sheet. Users can choose between low-density polyethylene (LDPE) and polycarbonate (PC) walls in the multi-tunnel structure; and between clear glass and diffuse glass in the Venlo structure. Different formulae were generated for each choice. Diffuse glass is estimated to consume an additional 10 kWh m2 of electricity in the production process, based on data from the glass manufacturer Groglass. The foundations and main path are made of concrete. Transport of the structure to the greenhouse by lorry and ship is included. 2.2. Auxiliary equipment This stage includes the manufacture of all the substrate and its transport to the greenhouse by lorry and ship. Users can choose between three types of substrate: perlite, rockwool and volcanic gravel. Additionally, the multi-tunnel spreadsheet calculates the amount of LDPE used to cover the benches. Because most of the electricity in a multi-tunnel greenhouse is consumed by the irrigation system, the input of electricity consumption is included in this stage. 2.3. Climate control system This includes several heating options: no heating, heating with gasoil or natural gas, and combined heat and power (CHP) with natural gas. When CHP is selected by the user, the total amount of natural gas consumption refers to the production of thermal energy and electricity. To calculate the amount of natural gas used to heat the greenhouse, an energy allocation of natural gas consumption was carried out using 64.4% natural gas to heat the greenhouse (Blonk et al., 2009). As the Venlo spreadsheet is designed for a glass greenhouse under Central European climate conditions where heating is always necessary because of the cold winters, a formula was developed to calculate the amount of steel in the heat distribution system. Therefore, steel in the heating distribution system was included and a statistical regression was established between the amount of steel and the greenhouse perimeter by analysing several greenhouses of different sizes. The total amount of electricity for the Venlo structure is also included in this stage. 2.4. Fertilisers For the environmental assessment of fertilisers, users are asked to enter the total amount of N, P2O5 and K2O on the questionnaire. To assist users in the calculation of the total amount of N, P2O5 and K2O, a fertiliser calculator was developed by Dr. Luca Incrocci (Dipartimento di Biologia delle Piante Agrarie, University of Pisa). The calculator is an EXCELÔ spreadsheet that can calculate the total amount of nutrients in a growing cycle (expressed in kg ha1) and nutrient use efficiency (NUE). The fertiliser stage includes emissions due to fertiliser production and use. Emissions to air and water because of application were estimated from available models for open field crops, as there is no specific model to calculate nitrogen emissions for protected crops. Emissions were calculated on the basis of total N applied: NH3eN and N2OeN as 3% and 1.25%,

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195

189

Fig. 1. Production system diagram, including stages and processes for a tomato crop in: (1a) a multi-tunnel greenhouse under Southern European climate conditions; (1b) a Venlo glass greenhouse under Central European climate conditions.

respectively, of total N applied (Brentrup et al., 2000); and NOxeN as 10% of N2OeN emissions (Audsley et al., 1997). Whether or not leached nitrates reach surface aquatic systems depends on a great number of variables, such as soil characteristics, groundwater aquifers and the presence of surface aquatic systems. We calculated potential NO3eN emissions at river basin scale as 30% of total N (Van Drecht et al., 2003) after discounting N emissions into the air and taking into account an uptake of 1.89 g N kg1 of tomatoes (Sonneveld, 2000).

2.6. Waste management The transport of greenhouse materials to a recycling plant, landfill, incinerator and compost plant are included based on the percentage of treatment for each material. Landfill and incinerator emissions are included. Emissions at the recycling and compost plants are not included, following the cut-off method used by Ekvall and Tillman (1997). 2.7. Transport

2.5. Pesticides The production of all the pesticides applied to the crop is included, as well as the machinery for their application. As there is considerable variability in active ingredient formulations, evaluation is simplified by distinguishing between the total amount of fungicides and insecticides. The production of pesticides made no relevant contribution to the total production system, but pesticides can have significant effects on ecosystem toxicity. Toxicity was not evaluated, as this is a complex topic beyond the scope of this study. Recent studies have demonstrated that, in terms of human toxicity, appropriate pesticide management has no dangerous consequences (Sevigné et al., 2012).

As described above, transport of the frame and substrate to the greenhouse is included, as is transport of waste materials from the greenhouse to treatment plants. There are many datasets on types of lorries in the Ecoinvent database and, for the present study, two lorry processes have been selected for the calculator: “lorry 3.5e 16t, fleet average” for transport of the frame, substrate and waste materials to the landfill and incineration plant: and “lorry 3.5e7.5t, EURO5” for transport of the waste materials to recycling and compost plants. Due to the variability of agricultural practices, users are asked to enter the life span of the greenhouse and the substrate. On the multi-tunnel spreadsheet, they are also asked for the life span of the plastic covering and walls. In this environmental calculator, the

190

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195

Table 1 Processes included in multi-tunnel and Venlo greenhouse spreadsheets. Topic

Input

Units

Reference values Multi-tunnel

Venlo

Crop

Crop Yield Density Stems per plant Growth period Type of substrate Substrate life span Bag volume Plants per bag Number of spans Span width Span length Roof vents: total greenhouse number Gutter height Ridge height Greenhouse walls Greenhouse frame life span Greenhouse roof covering life span Greenhouse walls life span Total greenhouse electricity consumption Water consumption Irrigation system Open/Closed N P2O5 K2O Fungicides Insecticides Heating Fuel Fuel consumption

Name kg m2 p m2 u p1 Weeks Name Years L u u m m u m m Material Years Years Years kWh m2 L m2 Type kg m2 kg m2 kg m2 kg m2 kg m2 Type None m3 m2

Tomato 16.5 1.23 2 52 Perlite 3 30 3 18 8 135 36 4.5 5.8 PC 15 3 15 0.641 474.8 Open 0.060 0.038 0.117 0.00285 0.00038 No heating NO 0.00

Tomato 56.5 1.25 2 52 Rockwool 1 14 3 25 8 200 2000 6 6.8 Clear glass 15 15 15 10 794.4 Closed 0.169 0.041 0.185 0.0007 0.0003 CHP Natural gas 64.7

Substrate

Structure data

Energy consumption Watering Fertilizers

Pesticides Heating

default life-span of the greenhouse is estimated to be 15 years, in accordance with the European Committee for Standardization (CEN, 2001), although most growers extend the life span even further. This environmental calculator is free and available on the EUPHOROS website (http://www.euphoros.wur.nl/UK/) in different languages: Dutch, English, Hungarian, Italian and Spanish.

category is included. The results of the production system by stage are provided on the Detailed Results sheet. Based on default data for the production system, absolute values are included in Tables 2 and 3, and the relative contributions of the stages to the production system are represented in Figs. S1 and S2 in the supplementary data. The reference results of the production system based on default data are described in more detail as follows.

3. Results 3.1. Tomato production in a multi-tunnel greenhouse The calculator provides results for the potential environmental impacts of the production system on the Total Results and Detailed Results sheets. Results are calculated in the life cycle impact assessment phase of the LCA and the aim is to associate the inventory created on the basis of user data with category indicators. Results are expressed by functional unit (1 tonne of tomatoes) and for each one of the environmental impact categories included in the calculator. Results are presented as absolute values and on graphs on two sheets to facilitate user comprehension. The total results of the production system, based on user data and default values, are shown on the Total Results sheet. A brief description of each impact

The structure, fertilisers and auxiliary equipment were the major contributors in all impact categories. The structure made the highest contributions in the impact categories abiotic depletion (50%), global warming (37%), photochemical oxidation (54%) and cumulative energy demand (50%), due to the large amount of steel in the frame and plastic in the covering and floor. The structure was a major burden in a multi-tunnel greenhouse with no heating, since there were few energy and input requirements apart from electricity consumption, substrate, fertilisers and water. Fertilisers were the main burden in acidification (39%), mainly because of ammonia

Table 2 Results provided by the environmental calculator using default data for a tomato crop in a multi-tunnel greenhouse by functional unit (1 tonne of tomatoes). Impact category ADP AAP EUP GWP POP CED Water

Abiotic depletion, kg Sb eq Air acidification, kg SO2 eq Eutrophication, kg PO3 4 eq Global warming, kg CO2 eq Photochemical oxidation, kg C2H4 Cumulative energy demand, MJ Water use, m3

Total

Structure

Auxiliary equipment

Climate control system

Fertilisers

Pesticides

Waste

1.295 0.942 0.501 202.1 0.033 3144 28.78

0.642 0.332 0.130 74.74 0.017 1560

0.475 0.222 0.083 62.19 0.010 1217

0.000 0.000 0.000 0.000 0.000 0.000

0.149 0.370 0.279 61.74 0.004 297.0

0.018 0.012 0.008 2.263 0.001 43.84

0.011 0.006 0.001 1.257 0.000 25.34

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195

191

Table 3 Results provided by the environmental calculator using default data for a tomato crop in Venlo glass greenhouse by functional unit (1 tonne of tomatoes). Impact category ADP AAP EUP GWP POP CED Water

Abiotic depletion, kg Sb eq Air acidification, kg SO2 eq Eutrophication, kg PO3 4 eq Global warming, kg CO2 eq Photochemical oxidation, kg C2H4 Cumulative energy demand, MJ Water use, m3

Total

Structure

Auxiliary equipment

Climate control system

Fertilisers

Pesticides

Waste

14.68 3.252 0.8550 1928 0.2150 30,860 14.06

0.298 0.282 0.087 47.19 0.013 713.1

0.074 0.063 0.014 8.482 0.003 162.6

14.20 2.624 0.698 1823 0.196 29,770

0.099 0.280 0.054 47.64 0.002 196.4

0.002 0.001 0.001 0.238 0.000 4.656

0.005 0.003 0.001 0.706 0.000 12.84

emissions into the air during their application, and also in eutrophication (56%), because of nitrate emissions into the water, since the greenhouse used an open-loop irrigation system. The auxiliary equipment makes significant contributions because of substrate and electricity consumption (each between 16% and 39%), depending on the impact category. The climate control system made no contributions, as there was no heating. The pesticide and waste stages made contributions to the total production system of less than 3%.

3.2. Tomato production in a Venlo glass greenhouse The climate control system is the main burden in all impact categories, with contributions of between 81% and 97% of the total production system. The reason for this environmental impact was the consumption of natural gas to heat the greenhouse. The structure is the second contributor in all impact categories, with percentages between 2% and 10%; metal and glass were the major burdens in this stage. Fertilisers made contributions representing between 0.6% and 8.6% of the total, auxiliary equipment contributed between 0.4% and 1.9% of the total, and pesticides and waste management contributed around 0%. Several considerations should be taken into account when calculating results in the simulation of production systems regarding the datasets on country electricity production, fertilisers, pesticides and means of transport.

3.4. Fertilisers evaluation The environmental performance of fertilisers is an important topic in agricultural production systems, because of the emissions released during their production and application. For simplification, the calculator gives results for the total amount of nutrients N, P2O5 and K2O. In order to gain insight into different approaches to the environmental performance of fertilisers, the contributions of specific fertilisers in a nutrient solution were calculated and compared with the contributions of the equivalent amount of nutrients N, P2O5 and K2O (Table 4). The total amount of nutrients was determined using the fertiliser calculator provided on the spreadsheet for this purpose. The performance of specific and generic fertilisers was compared in terms of total emissions released during production and application. Emissions from application were calculated as described in Section 2.4, except for ammonia emissions into air (NH3eN) by specific fertilisers. In this case, the Audsley model (1997) was used for calculation and ammonia emissions were 2% of

a

3.3. Electricity evaluation The contribution of electricity consumption can differ depending on the electricity production mix of the country where the production system is analysed. To simplify the calculator, only the European electricity production mix is used to calculate the environmental impact of electricity consumption on the spreadsheets. To determine the difference, the multi-tunnel greenhouse and the Venlo glass greenhouse were analysed using the electricity production mix of Spain and the Netherlands, respectively. Results showed differences in stage contributions, including electricity consumption (the auxiliary equipment in the multi-tunnel greenhouse and the climate control system in the Venlo glass greenhouse). Nevertheless, the order of importance of the stage contributions to the total production system remained the same. For example, in the acidification category, electricity consumption was 9% of the auxiliary equipment contribution in the multi-tunnel greenhouse when using the average electricity production in Europe, and 18% when using the Spanish production mix (Fig. 2). In the eutrophication category, the Venlo glass greenhouse showed a 29% contribution of electricity consumption in the climate system when using the average electricity production in Europe, and a 16% contribution when using the electricity production mix in the Netherlands (Fig. 3).

b

Fig. 2. Multi-tunnel greenhouse contribution to AAP impact category with electricity production mix in: (2a) Europe; (2b) Spain.

192

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195

3.5. Pesticides evaluation

a

Emissions from pesticides were a minor burden in the total production system and were caused by the production of insecticides and fungicides. A case was evaluated by comparing emissions from the production of fungicides with the emissions from the production of the same amount of active ingredients. The amount of 14.9 kg ha1of fungicides corresponded to the application of 2.38 kg ha1 of chlorothalonil, 0.40 kg ha1 of mancozeb and 12.1 kg ha1 of fungicides. The contributions of total specific fungicides were higher than those of the active ingredients (between 4.5% and 15.2% higher), depending on the impact category.

b

3.6. Transport evaluation Transport is a minor burden in the total production system. Nevertheless, the contribution can change, depending on the distance from origin to destination. Many types of road transport are available and, consequently, there is considerable variability in the emissions they release into the atmosphere. Therefore, the selection of one lorry over another can produce different results in the assessment. Only two kinds of lorry are included in the calculator database (Section 2.7). To gain insight into the variability in the environmental impact of lorries, several road transport processes in the Ecoinvent database were compared. A van was also included for the transport of small amounts of materials, produce and green biomass. The results in Table 5 show the high emission variability of the means of transport in impact categories, ranging from 119.3% to 198.2%.

Fig. 3. Venlo glass greenhouse contribution to EUP impact category with electricity production mix in: (3a) Europe; (3b) the Netherlands.

the total N applied for calcium nitrate and potassium nitrate, and 2.5% for nitric acid. Datasets for emissions from the production and application of specific fertilisers were taken from the Ecoinvent database (Ecoinvent, 2010); and datasets for emissions from the production of generic fertilisers were taken from the LCAFoods database (Nielsen et al., 2003). The results showed that emissions from the production of specific fertilisers made higher contributions in all impact categories than the production of generic fertilisers (1.4e5.7 times greater); and application emissions made lower contributions in the categories of acidification, eutrophication, global warming and photochemical oxidation (between 0.66% and 0.92%). The comparison of the performance of specific and generic fertilisers is shown in Fig. 4aed. Table 4 Specific fertilisers and their equivalence in total amount of generic fertilisers; and nitrogen emissions to air.

Specific fertilisers Calcium nitrate Nitric acid (56%) Phosphoric acid (72%) Potassium nitrate Potassium sulphate Generic fertilisers N P2O5 K2O a b

Audsley et al., 1997. Brentrup et al., 2000.

Amount

Nitrogen emissions to air

kg ha1

NH3eN

N2OeN

NOxeN

1261 908.1 646.3 1303 1357

2% of Na 2.5% of Na

1.25% of Nb 1.25% of Nb

10% of N2OeNa 10% of N2OeNa

2% of Na

1.25% of Nb

10% of N2OeNa

533.2 568.7 1312

3% of Nb

1.25% of Nb

10% of N2OeNa

4. Discussion In this section we discuss the main results of the study, the methodology used and the advantages and limitations of the environmental tool. Additionally, we propose several lines for future research to improve the calculator. The main objective of the study was achieved by developing an environmental software tool to calculate the environmental impacts of protected horticultural systems. The authors succeeded in providing a calculator with a simple design that can be used by all stakeholders in agricultural production systems as a support tool to take decisions and reduce the environmental impacts of greenhouse crops. All actors in agriculture, such as producers, advisers, consultants and politicians, have a certain degree of responsibility for environmental development and a tool like this one can be very useful in decision making. The LCA methodology was useful for the purpose and allowed for transparent and objective calculation of the environmental impacts of a production system throughout its life cycle. The tool was carefully simplified without compromising the principles of the ISO 14040 standard (ISO-14040, 2006) and while following the ILCD guidelines (ILCD, 2010) to justify the inclusion or exclusion of certain processes. Although it may seem that many elements of a full horticultural production system have been eliminated, the calculator includes a justified representation of the most significant ones. This calculator is expected to be very helpful to users to simulate the environmental performance of their own production systems and analyse alternatives of cleaner production, such as reducing the amount of fertilisers and energy consumption. The final results provided by the calculator are approximate and effectively serve to establish an initial situation to which alternatives can be compared and to help choose the most appropriate options. Three case studies were reported by Torrellas et al. (2012a). In the evaluation

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195

a

b

c

d

193

Fig. 4. Comparison between generic and specific fertiliser contributions to the following impact categories, based on emissions from production and application: (4a) acidification; (4b) eutrophication; (4c) global warming; (4d) photochemical oxidation.

framework, financial aspects should be included to fulfil the principle of development, as financial return is fundamental to all projects, a project may be environmentally sound, but too expensive to build (Ding, 2008). This tool does not include an economic appraisal, but an economic assessment of greenhouse crops can be carried out using a separate tool developed for this purpose and following cost-benefit analysis. The tool is available on the EUPHOROS website (EUPHOROS, 2008e2012). The environmental performance of a protected crop is represented by the inclusion of multiple environmental impact categories. This is a distinctive feature of this calculator, as most of the others on the market give only the contribution of greenhouse gas emissions to the category of global warming, which provides only a partial vision of the environmental problem (Finkbeiner, 2009). The inclusion of two spreadsheets helps distinguish the environmental evaluation of two types of protected crops: a plastic multi-tunnel and a Venlo glass greenhouse with different input requirements and, consequently, different environmental impacts. The calculation of environmental impacts is provided for by including appropriate formulae for two of the most significant aspects, the structure and heating, and calls for little data from the user on the greenhouse dimensions and energy consumption. In the case of CHP for heating the greenhouse, different approaches

and allocation factors could have been used, but we used an energy allocation, as it was useful to determine the amount of natural gas for heating and the amount of natural gas for electricity production could be excluded (Blonk et al., 2009; Torrellas et al., 2012b). This calculator is a contribution to help estimate the environmental impacts of agricultural production systems, aiming to reduce the complexity of environmental software tools targeted to the scientific community (Matthies et al., 2007; Schreinemachers and Berger, 2011). The calculator addresses specific issues of protected crops such as the structure, type of substrate and heating. Moreover, this tool gives insight into a range of environmental damages by providing results in six impact categories. It is different from other calculators on the market, such as the Cool Farm Tool (Hillier et al., 2011), which can calculate greenhouse gas emissions from crop and livestock production and takes into account the type of soil and machinery, but only for the global warming impact category. If simplicity is one of the advantages of the tool, the authors discovered that this was also the main cause of limitations. Agriculture involves a high variability of production systems because of such factors as geography, climate, soil characteristics, water availability and management practices that include conventional and organic farming (Haas et al., 2001). All these features could not

Table 5 Variability of eight types of road transport by impact category.

L16e32t, EU4 L16e32t, EU5 L 3.5e16t L 3.5e7.5t, EU4 L 3.5e7.5t, EU5 L 7,5e16, EU4 L 7,5e16, EU5 Van L16e32t, EU4 AVG CV

ADP kg Sb eq

AAP kg SO2 eq

EUP kg PO3 4 eq

GWP kg CO2 eq

POP kg C2H4

CED MJ

0.00117 0.00119 0.00183 0.00328 0.00331 0.00155 0.00157 0.01313 0.00117 0.00338 1.19321

0.00064 0.00050 0.00135 0.00169 0.00138 0.00083 0.00066 0.00725 0.00064 0.00179 1.25739

0.00017 0.00013 0.00036 0.00047 0.00039 0.00022 0.00017 0.00240 0.00017 0.00054 1.41613

0.16509 0.16730 0.25697 0.46726 0.47198 0.22093 0.22385 1.91035 0.16509 0.48547 1.21248

0.00002 0.00002 0.00004 0.00006 0.00006 0.00003 0.00003 0.00071 0.00002 0.00012 1.98164

2.76134 2.79363 4.46890 7.99663 8.06649 3.68501 3.72805 33.0266 2.76134 8.31583 1.22754

194

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195

be implemented in the calculator. Including the European electricity production mix means that the regional variation of electricity production is not taken into account and, consequently, more precise calculation of emissions is not possible. Nevertheless, for the two scenarios analysed, the order of importance of the stage contributions remained the same, despite the variation of electricity environmental impacts when country electricity production mixes were considered. This type of limitation could have been resolved by making the calculator an open source software for users, but this would have added complexity to users, which was one of the hazards to be avoided. Throughout the development of the calculator, the authors had to take decisions to achieve the aim of simplification and the desire for completeness. Management of fertiliser and pesticide application is an important issue in agriculture and therefore relevant in the agriculture LCA as well. Significant variation in results was observed when considering generic and specific fertilisers and pesticides. In the case of fertilisers, it would be useful to reach a consensus on the emission models of nutrients N, P2O5 and K2O and to have simplified generic datasets. For pesticides, even though their production was not found to be a major burden in the total production system, more datasets for specific fungicides and insecticides should be available. Pesticide toxicity is another important issue in agricultural systems, but it was not included in the calculator, as there is no general consensus for evaluation (Antón, 2008). Regarding transport, high variability was observed in the emissions released into the atmosphere, depending on the means of road transport. The transport contribution was low in the two scenarios analysed and, consequently, the selection of one means over another would not have changed the results significantly. Nevertheless, this is a point to be taken into account, as it could have a greater effect on production systems in which the transport contribution is higher. This environmental tool was finalised and subsequently tested by a group of agriculture support technicians. Nevertheless, the authors feel the tool could be improved by including the items discussed above. The results could be more detailed for fertilisers and pesticides, and an evaluation of pesticide toxicity could be included. Impact categories and characterisation models could be updated with models such as RéCiPe (Goedkoop et al., 2009) and USEtox (Rosenbaum et al., 2008). As in most LCA studies, this calculator includes water use quantification for the water consumption assessment while ignoring the environmental impact of water consumption on available freshwater resources, ecosystem quality and human health (Núñez et al., 2012). The water use assessment could be more accurate by including models that take into account the spatial differentiation of impacts at the watershed level (Milà i Canals et al., 2009; Pfister et al., 2009; Berger and Finkbeiner, 2010). Scenarios for more types of crops are desirable, as they would include more horticultural production system variability. If the tool used open source software, users could include their regional features such as electricity production, water availability and other kinds of substrate. And, of course, feedback from users would be very valuable to adapt the tool to their needs. Thus, this environmental tool is the starting point for future research and to implement developments achieved in other research areas, while always taking into account simplicity of use for final users. 5. Conclusions This is a useful tool to calculate the environmental impacts of protected crops and to support decision making. The tool is designed for a broad range of target users such as growers, advisers and consultants. The present study revealed that there is a lack of

simple high-quality data on fertilisers and pesticides. This tool could be improved by including more kinds of agricultural systems, site-specific variability for water use assessment and electricity production and by including user feedback. Moreover, this environmental tool could be the starting point for future research in different fields related to horticultural production systems. Acknowledgements This work is part of the EUPHOROS research project “Efficient use of inputs in protected horticulture” FP7-KBBE-2007-1 and financially supported by the European Union. The authors would like to thank all the partners involved in the project, especially Esteban José Baeza, Juan Carlos López, Marc Ruijs and Nieves García Victoria for providing data. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jenvman.2013.01.011. References Amores Barrero, M.J., Ferrer, F., Cabot, O., Duaigües, A., Antón, A., Castells, F., 2012. Implementing decision making in water management in agriculture based on environmental indicators. In: Proceedings of the Eighth International Conference on Life Cycle Assessment in the Agri-food Sector, Saint Malo, France. Antón, A., 2008. Sustainable management in horticulture: a life cycle perspective. In: Proceedings of the Eighth International Conference on EcoBalance, Tokyo, Japan. Athena EcoCalculator, July 2012. http://www.athenasmi.org/our-software-data/ ecocalculator/. Audsley, E., Alber, S., Clift, R., Cowell, S., Crettaz, P., Gaillard, G., Hausheer, J., Jolliet, O., Kleijn, R., Mortensen, B., Pearce, D., Roger, E., Teulon, H., Weidema, B., van Zeijts, H., 1997. Harmonisation of Environmental Life Cycle Assessment for Agriculture. European Commission DG VI Agriculture. Azapagic, A., Stichnothe, H., Gujiba, H., Morgan, A., Espinoza Orias, N., Rong, Y., Jeswani, H., Amienyo, D., 2010. CCaLC, Carbon Footprint Tool. July 2012. http:// www.ccalc.org.uk/. Berger, M., Finkbeiner, M., 2010. Water footprinting: how to address water use in life cycle assessment? Sustainability 2, 919e944. Blonk, H., Koll, A., Luske, B., Ponsioen, T., Scholten, J., 2009. Berekening van Verkenning en oplossingen van methodiekvragen ten behoeve van de ontwikkeling van het Nederlandse footprint. Blonk Milieu Advies BV, Gouda, the Netherlands. Brentrup, F., Küsters, J., Lammel, J., Kuchlmann, H., 2000. Methods to estimate onfield nitrogen emissions from crop production as an input to LCA studies in the agricultural sector. Int. J. Life Cycle Ass. 5 (6), 349e357. CEN, 2001. EN 13031. Greenhouses: Design and Construction e Part 1: Commercial Production Greenhouses. European Committee for Standardization. Ding, G., 2008. Sustainable constructiondthe role of environmental assessment tools. J. Environ. Manage. 86, 451e464. Ecoinvent, 2010. Ecoinvent Data v2.2. Swiss Centre for Life Cycle Inventories. Ekvall, T., Tillman, A., 1997. Open-loop recycling, criteria for allocation procedures. Int. J. Life Cycle Ass. 2 (3), 155e162. Epp, A., Reichenbach, A., 1999. ULME. Umweltfolgen von Lebensmitteleinkäufen. July 2012. http://www.ulme.ethz.ch/. EUPHOROS, 2008. Efficient Use of Inputs in Protected Horticulture. http://www. euphoros.wur.nl/UK/. Farreny, R., Gasol, C.M., 2012. CO2ZW: Carbon Footprint Tool for Waste Management in Europe. July 2012. http://co2zw.eu.sostenipra.cat/. Finkbeiner, M., 2009. Carbon footprintingdopportunities and threats. Int. J. Life Cycle Ass. 14, 91e94. Finnveden, G., Hauschild, M., Ekwall, T., Guinée, J., Heijungs, R., Hellweg, S., Kowhler, A., Pennington, D., Suh, S., 2009. Recent developments in life cycle assessment. J. Environ. Manage. 91, 1e21. Fundación_Cajamar, 2008. Análisis de la campaña Hortofrutícola de Almería, Campaña 2007/2008. Informes y Monografias, Almería. Goedkoop, M., Heijungs, R., Huijbregts, M., De Schryver, A., Struijs, J., van Zelm, R., 2009. ReCiPe 2008. A Life Cycle Impact Assessment Method Which Comprises Harmonised Category Indicators at the Midpoint and the Endpoint Level. PRé Consultants, Amersfoort, the Netherlands. Guinée, J.B., Gorrée, M., Heijungs, R., Huppes, G., Kleijn, R., de Koning, A., van Oers, L., Wegener Sleeswijk, A., Suh, S., Udo de Haes, H., Bruijn, H., van Duin, R., Huijbregts, M.A.J., 2002. Handbook on Life Cycle Assessment. Operational Guide to the ISO Standards. Kluwer Academic Publishers, Dordrecht, the Netherlands.

M. Torrellas et al. / Journal of Environmental Management 118 (2013) 186e195 Haas, G., Wetterich, F., Köpke, U., 2001. Comparing intensive, extensified and organic grassland farming in southern Germany by process life cycle assessment. Agr. Ecosyst. Environ. 83 (1e2), 43e53. Halberg, N., 2004a. DIAS report. In: Proceedings of the Fourth International Conference on Life Cycle Assessment in the Agri-food Sector. Danish Institute of Agricultural Sciences, Bygholm, Denmark. Halberg, N., 2004b. How may quality assurance systems in food chains include environmental aspects based on life cycle methodology?. In: Proceedings of the Fourth International Conference on Life Cycle Assessment in the Agri-food Sector. Bygholm, Denmark. Hillier, J., Water, C., Malin, D., Garcia-Suarez, T., Mila-i-Canals, L., Smith, P., 2011. A farm-focused calculator for emissions from crop and livestock production. Environ. Modell. Softw. 26, 1070e1078. idc LCA calculator, July 2012. http://www.lcacalculator.com/. ILCD, 2010. International Reference Life Cycle Data System (ILCD) Handbook. General Guide for Life Cycle Assessment e Detailed Guidance. European Commission. Joint Research Centre. Institute for Environment and Sustainability. ISO-14040, 2006. Environmental Management e Life Cycle Assessment e Principles and Framework. International Organisation for Standardisation ISO, Geneva, Switzerland. Kellenberger, D., 2010. Web-based environmental and economic impact calculator for typical NZ residential wall components. In: Proceedings of the Tenth New Zealand Sustainable Building Conference, Wellington, New Zealand. LiMaS WebSuite, July 2012. http://limas.simpple.com/limas-websuite/. Matthies, M., Giupponi, C., Ostendorf, B., 2007. Environmental decision support systems: current issues, methods and tools. Environ. Modell. Softw. 22 (2), 123e127. Milà i Canals, L., Chenoweth, J., Chapagain, A., Orr, S., Antón, A., Clift, R., 2009. Assessing freshwater use impacts in LCA: part I-inventory modelling and characterisation factors for the main impact pathways. Int. J. Life Cycle Ass. 14, 28e42. Montero, J.I., Antón, A., Torrellas, M., Ruijs, M., Vermeulen, P., 2011. Report on Environmental and Economic Profile of Present Greenhouse Production Systems in Europe. European Commission. Mourad, A.L., Coltro, L., Oliveira, P.A.P.L.V., 2007. A simple methodology for elaborating the life cycle inventory of agricultural products. Int. J. Life Cycle Ass. 12 (6), 408e413. Muñoz, P., Antón, A., Paranjpe, A., Ariño, J., Montero, J.I., 2008. High decrease in nitrate leaching by lower N input without reducing greenhouse tomato yield. Agron. Sustain Dev. 28, 489e495. Nielsen, P., Nielsen, A., Weidema, B., Dalgaard, R., Halberg, N., 2003. LCA Food Data Base. http://www.lcafood.dk. Núñez, M., Pfister, S., Antón, A., Muñoz, P., Hellweg, S., Koehler, A., Rieradevall, J., 2012. Assessing the environmental impact of water consumption by energy

195

crops grown in Spain. J. Ind. Ecol.. http://dx.doi.org/10.1111/j.15309290.2011.00449.x Okala Life Cycle Analysis Calculator, July 2012. http://www.ridspedia.net/lca/. Pascualino, J., Meneses, M., Castells, F., 2009. JocAPQUA: Joc de la compra sostenible. July 2012. http://jocapqua.urv.es/. Pfister, S., Koehler, A., Hellweg, S., 2009. Assessing the environmental impacts of freshwater consumption in LCA. Environ. Sci. Technol. 43 (11), 4098e4104. Pidd, M., 1996. Five simple principles of modelling. In: Proceedings of the 1996 Winter Simulation Conference, Coronado, California, USA. Pires, A., Martinho, G., Chang, N., 2011. Solid waste management in European countries: a review of systems analysis techniques. J. Environ. Manage. 92, 1033e1050. Robust DSC LCA Calculator, July 2012. 3GSolar. http://www.3gsolar.com/?page_ id¼47. Rosenbaum, R.K., Bachmann, T.M., Swirsky Gold, L., Huijbregts, M.A.J., Jolliet, O., Juraske, R., Koehler, A., Larsen, H.F., MacLeod, M., Margni, M., McKone, T.E., Payet, J., Schuhmacher, M., van de Meent, D., Hauschild, M.Z., 2008. USEtoxdthe UNEP-SETAC toxicity model: recommended characterisation factors for human toxicity and freshwater ecotoxicity in life cycle impact assessment. Int. J. Life Cycle Ass. 13, 532e546. Schreinemachers, P., Berger, T., 2011. An agent-based simulation model of human environment interactions in agricultural systems. Environ. Model. Softw. 26, 845e859. Sevigné, E., Fantke, P., Juraske, R., Kounina, A., Antón, A., 2012. Deposition and residues of azoxystrobin and imidacloprid on greenhouse lettuce with implications for human consumption. Chemosphere 89 (9), 1034e1041. Sonneveld, C., 2000. Effects of Salinity on Substrate Grown Vegetables and Ornamentals in Greenhouse Horticulture. Wageningen University, Wageningen, Ph.D thesis. Stoate, C., Báldi, A., Beja, P., Boatman, N.D., Herzon, I., van Doorn, A., Snoo, G.R., Rakosy, L., Ramwell, C., 2009. Ecological impacts of early 21st century agricultural change in Europe e a review. J. Environ. Manage. 91, 22e46. Torrellas, M., Antón, A., Montero, J.I., 2012a. An environmental impact calculator for greenhouse production systems. In: Proceedings of the Eighth International Conference on Life Cycle Assessment in the Agri-food Sector, Saint Malo, France. Torrellas, M., Antón, A., Ruijs, M., García, N., Stanghellini, C., Montero, J.I., 2012b. Environmental and economic assessment of protected crops in four European scenarios. J. Clean. Prod. 28, 45e55. Van Drecht, G., Bouwman, A.F., Knoop, J.M., Beusen, A.H.W., Meinardi, C.R., 2003. Global modeling of the fate of nitrogen from point and nonpoint sources in soils, groundwater, and surface water. Glob. Biogeochem. Cy. 17 (4), 1115e1134. Vermeulen, P.C.M., 2008. Quantitative Information for the Greenhouse Horticulture; Greenhouse Vegetables e Cut Flowers e Potted Plants. Report 185. Wageningen UR Greenhouse Horticulture, Bleiswijk, the Netherlands.