Renewable energy derived from agricultural biomass in peripheral rural areas: ‘Vicious circle,’ ‘Gordian knots,’ and turning points

Highlights

• A holistic methodology of research on peripheral rural areas.

• ‘Vicious circle’ of underutilization of the renewable energy of agricultural biomass.

• Propensity to grow energy crops and use agricultural biomass as renewable energy.

• ‘Gordian knots’ of omissions and public policies errors .

• •‘Alexandrian cut,’ or a bold action that severs the ‘Gordian knots’ decisively.

Abstract

The holistic methodology of the article consists of linking spatially differentiated cumulative causality processes and related structural issues in the ‘vicious circle’ of the underutilization of the renewable energy potential of agricultural biomass, with the ‘Gordian knots’ of omissions and errors in public policies. A Gordian knot is a metaphor for a cluster of interdependent, multidimensional, and excessively complicated problems that cannot be solved in a conventional way that can only be ‘disentangled’ by an ‘Alexandrian cut,’ or a bold action that severs it decisively. The conceptual framework matrix integrates the turning points approach and applies it to the interrelationship between internally breaking down the underutilization of the renewable energy potential of agricultural biomass and Alexandrian cuts that create the conditions for a bottom-up uptake of energy crops and renewable energy derived from agricultural biomass in peripheral rural areas. This study relies on a survey conducted on a sample of 2,191 farmers as well as secondary data to contextualize the survey data. We indicate that an interaction of cumulative causality processes and related structural issues ‘tighten’ the vicious circle, making reciprocally Gordian knots ‘impossible to untangle’.

Introduction

Research on sustainable energy often concerns political will, regulations and policies (Alexander et al., 2015), as well as funding, support, and the financial-economic performance of renewable energy projects. A project can be deemed successful when it generates social benefits and positive environmental impact (Feola and Nunes, 2014), and when it continues to operate in the long term. When analyzing the determinants of success of renewable energy projects that use agricultural biomass, researchers focus on leadership and intermediaries, human capital and tacit knowledge as well as institutional capacity building in civil society and collective learning embedded in social networks. When a core group of farmers who are best acquainted with local conditions participate in the collaborative planning and management of renewable energy projects that use agricultural biomass (or, more broadly, sustainable development), they come to see them as their own and identify with them, which ultimately becomes a source of trust, fairness, and social ‘license’ for such projects (Lang et al., 2012). Social acceptance is closely linked to the knowledge and environmental values of residents as well as the visual attractiveness of projects. Social acceptance, involvement of local actors, and successful implementation all contribute to positive perceptions of renewable energy projects (Gölz and Wedderhoff, 2018).

Energy projects that use agricultural biomass encounter opposition when they benefit individuals or specific stakeholder groups at the expense of other individuals and the entire community. However, policymakers often fail to recognize the social, economic, environmental, and spatial conflicts related to energy derived from agricultural biomass. Geels et al., (2017) points to the need for multi-level research on renewable energy, comprising analyses of diverse interests, values, knowledge, and resources among residents, farmers, and local and regional authorities as well as national governments, whose policies should be studied in depth. Rossi and Hinrichs (2011) indicate that public authorities responsible for energy policy often view farmers instrumentally as suppliers of agricultural biomass devoid of their own opinions on renewable energy. Bayulgen and Benegal (2019) as well as Beer and Theuvsen (2019) emphasize that little research has been done to date on farmers' knowledge of the use of renewable energy derived from agricultural biomass. Our study rejects the assumption of neoclassical economics that farmers have full knowledge, analyze the costs and benefits of alternative actions, and make rational decisions. Instead, our hypothesis is that farmers’ knowledge about energy crops and the use of agricultural biomass for energy purposes is fragmented.

De Wit et al. (2011) point to agricultural policy and progress in technology as the key factors to take full advantage of the potential of agricultural lands to produce energy crops. However, the uptake of energy crops 'and renewable energy derived from agricultural biomass is determined by a combination of interlinked, multidimensional political, institutional, cultural, and economic processes, not just technological ones (Soares da Silva and Horlings 2019). Warbroek et al. (2019) emphasize that the literature lacks both a holistic diagnosis and a holistic approach to solving the problem of how and why the potential of low-carbon energy in rural areas is underutilized. The assumption of the ‘logical architecture’ (methodology) of our work is that transdisciplinarity — i.e. transcending the boundaries of conventional knowledge in narrow scientific disciplines — is a key element of sustainability, as shown by Brandt et al. (2013) and other authors. The integration of knowledge and practice from different scientific disciplines is essential to achieving the overarching objectives of renewable energy projects that are important in the long term for communities and collectives (Lang et al., 2012). As noted by Kliskey et al. (2017) transdisciplinarity is insufficiently reflected in sustainability science.

This paper fills a gap in the literature on identifying and classifying the interactions between the objective factors underlying the production of agricultural biomass for energy purposes in rural areas and the subjective factors, which are reflected in the declarations, opinions, and knowledge of farmers. The classification of farmers developed here according to their propensity to grow energy crops and use agricultural biomass as a source of renewable energy (taking into account the characteristics of their farms) contributes to scientific knowledge, which is, in a way, co-created by local actors. The holistic, strategic approach, presented in our study, which consists of linking spatially differentiated cumulative processes — i.e. multi-level combinations of objective and subjective factors with omissions and errors in public policy, as well as overcoming internal barriers to the uptake of renewable energy derived from agricultural biomass in rural areas and solving the associated ‘Gordian knots’ of omissions and public policy errors through ‘Alexandrian cuts,’ contributes to the existing methodology of sustainability science. This study makes a contribution to sustainable policy reforms to improve the implementation of United Nations Sustainable Development Goal of providing affordable, reliable, sustainable and modern energy for all as well as inclusive economic growth and reduced inequality.

According to Hudson at al. (2019) policy implementation creates a system of multidimensional, interconnected processes, differentiated in space and time, and therefore influenced by regional and local factors. Clarifying the role of socio-geographic site specificity, ‘spatialities’, and place-based conditions in the uptake of energy crops and renewable energy derived from agricultural biomass in rural areas as well as in making public policies connected with the transition, remains underexplored and potentially undervalued (Naumann and Rudolph, 2020, Süsser et al., 2017; Bridge 2013; Devine-Wright 2011). Gölz and Wedderhoff (2018) emphasize that research on renewable energy often involves regional or local analysis, but the local territorial units of different regions are characterized by different combinations of interconnected conditions — and thus, widely varying renewable energy potentials (Scaramuzzino et al., 2019; Süsser et al., 2017). Researchers stress the need for research on a subregional level (Balest et al., 2019), arguing that the integration of resources and coordination of local projects enables economies of scale and synergies in areas larger than a single municipality. Our paper contributes to this body of literature by developing classifications of supralocal territorial units according to objective and subjective determinants as well as the diversity of renewable energy potentials of agricultural biomass.

The analysis was conducted on the example of peripheral rural areas of Poland, a country in transition where coal is the primary source of energy. The analyzed areas belong to the Lublin region, which has a population of more than 2.1 million and low urbanization, with cities accounting for 46.5% of its population (https://bdl.stat.gov.pl/BDL/dane/teryt/jednostka. Accessed Nov 7, 2019). Its peripheral geographical status is usually operationalized as a combination of its location on the outer edges of both the European Union and Poland, poor transport accessibility, and low GDP per capita — 47% of the EU average in Purchasing Power Standard (https://bdl.stat.gov.pl/BDL/dane/teryt/jednostka. Accessed Nov 7, 2019). However, we understand its peripheralization as an interaction of the ‘vicious circle’ of underutilization of the region's potential, and related structural omissions and errors in public policies. The Lublin region is interesting for international researchers and practitioners because, like Eastern and Southern Europe, it can be described as an area where unfavorable phenomena and processes accumulate (Jakubowski, 2017), as do policy omissions and errors at the supranational (EU), national, regional, and local levels, all of which form their own ‘Gordian knots.’ A ‘Gordian knot’ is a metaphor for a cluster of interdependent, multidimensional, and excessively complicated problems that cannot be solved in a conventional way that can only be ‘disentangled’ by an ‘Alexandrian cut’, or a bold action that severs it decisively. Gordian knots can be operationalized as accumulations of unformulated, uninitiated, or unrealized strategic decisions which, if made and implemented, would create conditions for changes in the economy and society of the peripheral rural region from within (Kukliński, 2010).

The cultivation of energy crops on fallow land and the use of renewable energy, in addition to tourism and agricultural processing, are among the most commonly noted directions of development for rural areas in peripheral regions (Bański and Mazur, 2016). Roddis et al. (2019) highlight the complexity of the underlying factors and the need to research the potential of renewable energy as a basis for sustainable energy transition policies. In response to this, our study develops a method for estimating the potential of renewable energy derived from agricultural biomass, and identifies both its mass and spatial heterogeneity. The hypothesis is that the rural areas of the Lublin region are characterized by a high agricultural biomass potential, but this biomass is either used for other purposes than the production of renewable energy or left undeveloped.

The article begins with a theoretical background and conceptual framework that integrate two key concepts— cumulative causality processes and Gordian knots— with the turning points of overcoming the underutilization of the renewable energy potential of agricultural biomass. Then, we present the interplay of negative cumulative processes and related structural issues in the vicious circle of the peripheral rural region, and their interdependencies with the Gordian knots of omissions and public policy errors. The ‘last section’ of the article comprises a discussion of the results and formulates conclusions as well as recommendations of Alexander cuts for policymakers to take full advantage of the renewable energy potential of agricultural biomass.

In the long run, external benefits accumulate in the core areas due to the spatial concentration of capital and the diversity of economic structures, knowledge and human capital, and infrastructural facilities. Economies of scale are linked to the cumulative strength of the externalities as pull factor of economic activities (Krugman, 1991). External benefits include the positive economic and non-economic effects that arise as a result of external influences. Non-economic benefits are often institutional in nature and encompass both formal rules (e.g., laws, regulations) and informal norms (e.g., trust and fairness, solidarity, co-responsibility), which are often more significant than their formal counterparts (North, 2003). In core areas, institutions are often tasked with driving innovation and developmental impacts, while in peripheral rural areas they are often expectational and populist (Farole et al., 2011). In core areas, externalities include the benefits of acquiring tacit knowledge through direct contacts, which often entails low transport costs. Core areas attract tangible and intangible resources, especially human capital, draining peripheral rural areas that accumulate negative externalities and move away from equilibrium, which runs counter to the stable equilibrium approach (Panico and Rizza, 2009). According to Myrdal (1957) their development depends on the cores from which direct investment as well as infrastructural investment and social transfers (as public policy interventions) flow to the periphery. The political, economic, and social structures of these hinterlands may therefore have a vested interest in perpetuating the vicious circle of backwardness. Public authorities at the central, regional, and local level often lack the capacity to formulate and implement long-term, sequential plans of action aimed at curbing the processes of cumulative destruction, or to critically assess the omissions and errors of public policies (Gorzelak, 2015), which leads to the creation of an ever-increasing number of Gordian knots (Kukliński, 2010; Röpke, 2007).

Peripheralization can be understood as a set of processes that create peripheries through social, economic and political polarization, and in the spatial dimension, amplify discrepancies and variations on socioeconomic variables (Kühn, 2015). Munro (2019) conceptualizes peripheralization as a combination of multi-dimensional processes of creating peripheries within a socio-technical transition, as well as their interdependencies set in a framework of wider contexts. We define peripheralization as an interaction of cumulative processes and related structural issues ‘tightening’ the vicious circle of underutilization of the regional and local potentials that reciprocally ‘tighten’ the Gordian knots of omissions and errors in public policies, making the knots impossible to disentangle. The holistic methodology of our study identifies the dependencies and interactions among objective and subjective cumulative causality processes, namely the vicious circle (particularly in relation to renewable energy derived from agricultural biomass) as well as the Gordian knots (Fig. 1). It also encompasses strategic conceptual and practical approaches to address the problem, which Jerneck et al. (2011) and Lang et al. (2012) consider a key aspect of the methodology of sustainability science. The conceptual framework matrix therefore incorporates distinct turning points, i.e. the interdependence between breaking down barriers to the underutilization of the peripheral rural area's potential from within and Alexandrian cuts performed on existing Gordian knots – that is, unconventional activities and public policies that may conflict with established knowledge and practice in the main currents of thought in a given scientific discipline or disciplines. ‘The metaphorical severing’ of the Gordian knots does not mean that positive structural changes will automatically take place in the peripheral rural region, as those are generated not by one-off acts, but by consistent long-term strategic and operational measures, using dedicated instruments to implement European, national, regional, and local policies.

A survey was conducted among farmers in 2011–12 (today, under less favorable conditions than, the willingness of farmers to grow energy crops would probably turn out to be very low). The survey, preceded by pilot studies, consisted of 40 questions. The target area of the research comprised 111 gminy (roughly equivalent to communes) of the Lublin region, which were selected through stratified random sampling. University students conducting fieldwork exercises in socioeconomic geography assisted in administering the surveys used in the study. In-depth interviews were also conducted with individuals responsible for administering the survey, farmers, and experts. The analysis included the results of 2,191 questionnaires, which covered 1.15% of farms with an area greater than 1 ha (0.85% of all farms) and 1.89% of the agricultural land in the Lublin region (Table 1). Most of the respondents were farmers with holdings of 5 ha or more. The average area of the surveyed farm was 14.3 ha, which constituted 169% of the average for the Lublin region.

The representativeness of the sample was tested using p-value metrics (the spatial autocorrelation Moran's I test demonstrates a very low, statistically insignificant autocorrelation, i.e. the selection of communes for the study was random) as well as by comparing the area of croplands, meadows, and orchards in the Lublin region estimated in the questionnaires (using the $A_c$ formula) with Central Statistical Office in Warsaw (GUS) data.

$$A_c=\frac{A_{agr}}{A_{ags}}\sum _{i=1}^n{A}_i$$

• $A_c$ – estimated area of cultivated land in the region,

• $A_{agr}$ – area of agricultural lands in the region (Użytkowanie …, 2011, p. 46),

• $A_{ags}$ – area of agricultural holdings surveyed,

• $n$ – number of surveys conducted,

• $A_i$ – cultivated area for the i-th crop.

The results of the comparison indicate differences in the estimated area of meadows (which the respondents often omitted in their answers).

We used the point method to evaluate farmers’ propensity to grow energy crops and use agricultural biomass as a source of renewable energy. The analysis included responses to 17 questions: forms of fertilization, straw plowing (for compost), knowledge about energy crops, usage of land under these crops, and use of agricultural biomass as a source of renewable energy. We estimated 17 averages for responses to the questions (for each of the characteristics of farmers and their holdings – Table 3). Averages for favorable responses (indicating a willingness to grow energy crops and use renewable energy derived from agricultural biomass) that significantly deviated from the average for the region (by more than 5 percentage points) were given 1 point, while averages for unfavorable responses were given −1 point. Averages with very large deviations from the average for the region (by more than 10 p.p.) were given 2 or -2 points respectively. On this basis, a classification of farmers (their characteristics and the characteristics of their holdings) according to their propensity to grow energy crops and use renewable energy derived from agricultural biomass was developed (Table 3).

The process for classifying supralocal territorial units (powiaty, counties) comprised: aggregating and standardizing the data, removing variables with a variance below 10% and correlated at p < 0.05, simulating the classification, selecting the number of classes using Ward's method, and classifying using k-means clustering. The structural issues highlighted in the classifications originate from the survey data as well as secondary data drawn from publicly available statistical sources, which help to contextualize the survey data. The study made use of the GUS Local Data Bank and the results of the General Agricultural Census. The counties were classified according to objective factors influencing agricultural biomass production as a source of renewable energy, taking into account the yield of the marketable agricultural output (correlated with the quality of soils), the average area of an agricultural holding, population decline, and the attractiveness of counties as sites for agricultural biomass energy facilities; according to subjective factors, they were classified on the basis of respondents' answers, which concerned: the use of straw plowing for compost each year, declared sales of fixed agricultural byproducts (FAB), declared ownership and use of land for the cultivation of energy crops, as well as opinions on incentives to undertake their cultivation. In the synthetic classification of counties according to the renewable energy potential of agricultural biomass, the following variables were used: theoretical potential of fixed agricultural byproducts, technical potentials (FAB_tech1 and FAB_tech2), and the potential of energy crops according to the farmers surveyed (calculated on the basis of declared agricultural land for the cultivation of energy crops and the yield of biomass per hectare).

We first estimated the theoretical potential (mass) of FAB using the following formula:

$$FA{B}_{theor}=\frac{A_u}{A_r}\sum _{i=1}^n\sum _{k\in C_i}{A}_{kc}\cdot Y_{kc}\cdot W_{kc}\cdot W_{qaps}+\frac{A_u}{A_r}\sum _{i=1}^n\sum _{l\in S_i}{A}_l\cdot B{P}_l$$

• $FA{B}_{theor}$ - theoretical potential of fixed agricultural byproducts,

• $A_u$ - area of agricultural land in the territorial unit,

• $A_r$ – area of farms owned by respondents (farmers),

• $n$ - number of interviews conducted,

• $A_{kc}$ – area of cultivation of the k-th cereal, buckwheat, or rape,

• $C_i$ – set of cereal, buckwheat, or rape for the i-th farmer,

• $Y_{kc}$ - yield per hectare of the k-th cereal, buckwheat, or rape,

• $W_{kc}$ - ratio of the yield of straw to the yield of grain of the k-th cereal, buckwheat, or rape,

• $W_{qaps}$ - coefficient of quality of the agricultural production space in the commune,

• $A_l$ – cultivated area for the l-th crop,

• $S_i$ – set of crops for the i-th farmer,

• $B{P}_l$ - weight of byproducts from the cultivation of 1 ha the l-th crop.

The technical potential (FAB_tech1) was calculated on the basis of declarations by the surveyed farmers, which concerned the predominant type of fertilization (mineral or organic) and the application and frequency of straw plowing, according to the formula:

$$FA{B}_{tech1}=\sum _{i=1}^{n}FA{B}_{theor}\cdot F{T}_i\cdot S{P}_i$$

• $FA{B}_{tech1}$ - technical 1 potential of FAB,

• $FA{B}_{theor}$ - theoretical potential of FAB,

• $FT$ - fertilization type coefficient (Table 2),

• $SP$ - frequency of straw plowing coefficient (Table 2),

• $n$ - number of observations.

The potential of FAB_tech2 was estimated according to declared surpluses of straw, hay, and other harvest residues using the following formula:

$$FA{B}_{tech2}=\sum _{i=1}^n\sum _{j=1}^{c}FA{B}_{theor}\cdot B{T}_{ij}\cdot W{S}_{ij}$$

• $FA{B}_{tech2}$ - technical 2 potential of FAB,

• $FA{B}_{theor}$ - theoretical potential of FAB,

• $BT$ - coefficient for type of byproduct (Table 2),

• $WS$ - willingness to sell (portion of the surplus that the farmer aims to sell (Table 2),

• $n$ - number of observations.

• $c$ - number of byproducts.

The Euclidean distance between territorial units was calculated using Ward's method (Reddy and Vinzamuri, 2014), according to the formula:

$$d\left(G,H\right)=\frac{n_G{n}_H}{n_G+n_H}{\left|\right|m_G-m_H\left|\right|}^2$$

where ${\text{m}}_{\text{G}}$ i ${\text{m}}_{\text{H}}$ are the cluster centroids G and H a ${\text{n}}_{\text{G}}$ i ${\text{n}}_{\text{H}}$ represent the number of observations belonging to clusters G and H respectively.

Ward's method indicates the extent to which the sum of squared errors of prediction (SSE) increases when clusters G and H are combined into one. The SSE is a measure of the discrepancy between empirical values (${\text{y}}_{\text{i}}$) and theoretical predictions ($\widehat{{\text{y}}_{\text{i}}}$):

$$SSE=\sum _{i=1}^n{\left(y_i-\widehat{y_i}\right)}^2$$

The K-means algorithm minimizes the expression:

$$\sum _{i=1}^{n}d\left(x_i,m_{C\left(i\right)}\right)$$

which is the sum of the distance between the i-th observation and the centroid to which the observation belongs (Reddy and Vinzamuri, 2014).