APPLICATION OF GIS IN AGRO-RESIDUE BIOENERGY PLANNING

The  whole supply chain of a bioenergy project can  be  divided into three major spatially interlinked elements: (i) resource assess- ment, (ii) logistic planning, and (iii) power plant design. GIS inter- vention in  bioenergy planning is  necessary because: (i)  diverse varieties of agro-residues are  used as  energy feedstock, therefore maintaining their spatio-temporal database concerning physico- chemical characteristics, availability and distribution is important. GIS helps in  managing such database which later can  guide the industries for  effective collection of  raw  material,  allocation of the benefits of  bioenergy and cost-benefits analysis (Long  et al.,2013; Alfonso et al., 2009). Periodic updating of the biomass inven- tories is also  necessary to assess future feedstock supply potential; (ii) ensuring sustained feedstock supply is critical for  viable com- missioning of  a  power plant. Prior  knowledge of  any  fluctuation in   feedstock supply would  allow the  user  to   make  necessary arrangement for  alternative feedstock supply during lean period of  supply and it  can   be  indicated through GIS (Stephen et  al., 2010); (iii) consideration of the environmental requirement of resi- due and harvest constraints, economy, technology, competing uses of residue, local  socio-political dynamic, land use,  logistic facilities, civil  and industrial users can  also  be  assessed with GIS (Alfonso et al.,  2009; Beccali  et al.,  2009; Gomez et al.,  2010; Jiang  et al., 2012).

The  uneven geographical distribution of agro-residue demands proper logistic planning for the collection and transfer of residue in a time and cost-efficient manner. The  various parties involved in the biomass supply chain that influence the final  bioenergy cost include the supplier of  biomass, transportation and distribution entities, energy production facility developers and operators, gov- ernment and utility firms and the end users (Mafakheri and Nasiri, 2014). The  implementation of a spatial network can  help the par- ties  involved in bioenergy planning to act  in a common framework by sharing costs, logistics, and personnel (Beccali et al., 2009). The rising competition for  production areas, raw materials and infras- tructure also  demands spatially explicit logistic planning (Fiedler et al., 2007).

Among the  logistic parameters,  transportation   can   be   cost intensive depending upon the distance and mode of transport (rail- way,  roadway, waterway), nature of the feedstock (loose or dense) and the condition of the transport route. Optimization of the trans- portation network can  significantly reduce the cost  of  transport. GIS  has   the  capability to   model least-cost transport  pathway (Perpina et  al.,  2009; Ebadian et  al.,  2011; Jiang   et al.,  2012; Hiloidhari et al.,  2012). For  example, Network Analyst extension of  ArcGIS  software can   model least-cost transport  pathway  for delivering biomass feedstock from source to user location. It allows one  to  perform multiple network-based spatial functions such as identification  of  optimal/shortest   route,  closest  facility, service area, origin-destination  analysis (Perpina et  al.,  2009). Further, transporting  compacted  (baled) biomass reduces  cost   and CO2 emission, even for  short distances by  allowing higher amount to be  carried (Alfonso et al., 2009).

Another logistic parameter that requires managerial attention is the selection of optimal biomass collection area and power plant site,  where multiple factors come into play.  Selecting optimal bio- mass collection areas should be  in  accordance with existing agri- cultural,  geographical and  infrastructural  characteristics  of  the area  encompassing  supply  and  user  locations  (Beccali et  al., 2009). The site  should be easily accessible by transport route, near to utility points, feasible for the optimum planning of power trans- mission lines (Zubaryeva et al., 2012). Further, the plant should be installed  in   reasonable distance from residential areas, nature reserves to minimize potential negative impacts of plant operation and waste disposal. Beccali  et al. (2009) developed a GIS method to assess the techno-economic potential of biomass for energy gener- ation in Sicily through identification of efficient transportation net- work and optimal biomass collection areas. Similarly, Fiedler et al. (2007) designed a GIS logistic model for cost  efficient supply of bio- mass feedstock to industrial units by analyzing the profitability of investment in infrastructure and equipment for biomass supply.

GIS  based biomass power  plant  site   selection can   be   done through two methods: (i) suitability analysis (ii) optimally analysis as elaborated by Shi et al. (2008). Suitability analysis allows user to identify the most suitable site  for a power plant among many can- didate sites, based on  user defined constrain and supportive crite- ria.    On    the   other  hand,  optimally  analysis  considers  the relationship between biomass and power plants in  order to  find the optimal power plant locations at minimum transportation cost. The  optimization of biomass power plant location could be  done either through location-allocation modelling or  supply-area mod- elling. Using the optimally analysis approach, Shi et al. (2008) iden- tified potential  sites  for   biomass  power  plants  in   Guangdong Province, China,  considering the cost  of transportation as  a prime determining factor in developing bioenergy plant.

Logistic   planning for  biomass collection can  be  further influ- enced by the type of land holding. For example, in India and China, due to  small agricultural  landholding by  the  farmer, collecting agro-residue biomass from farmland may be  complicated com- pared to the Western countries (Yu et al., 2012). For the same col- lection radius, the collectors have to  deal with a larger number of farmers,  with  organized  logistic of  contracting, collection and transportation (Yu et al., 2012). In a long  supply chain, distributed biomass receiving stations (satellite storage) must be optimized for least cost  delivery of biomass as reported by Yu et al. (2012). How- ever,  Gomez et al. (2010) reported that, the size  of a collection area produces two counteracting effects on the final  cost  of energy gen- erated. A large biomass collection area results in  higher power plant  capacity, making the  power  plant  economically reliable, but, it  increases the transportation  costs of  the biomass to  the plant. Therefore, the authors (Gomez et al., 2010) have suggested that, in a large scale study with large geographical biomass supply area (e.g.  country, province), it is impractical to  optimize the size and location of  every possible plant. But  the better option is  to use  the same collection area for the whole of the territory and size of this area is  to  be  determined in  a  way  to  fulfil  the individual plant’s minimum installed capacity (Gomez et al., 2010).

Egbendewe-Mondzozo et al. (2011) proposed a spatial bioeco- nomic model, an  integration of biophysical -GIS- economic regio- nal  mathematical optimization model to  estimate biomass supply from cellulosic crops and crop residues. The  GIS part provides the transportation parameters to  the bioeconomic model. Overall, the model can  predict how biomass supply and environmental conse- quences respond to changes in genetic and biological management as well  as market prices and government policy. Similarly, J. Singh et al. (2011) developed a GIS model for agro-residue based decen- tralized biomass power plant design at development block level  in Punjab, India,  which can  be useful to decide optimum power plant locations with minimum storage and handling cost  of feedstocks. GIS application has  also  been found suitable for village level  small scale energy planning (Hiloidhari and Baruah, 2011; Kaundinya et al., 2013). Table  1 summarizes selected numbers of recent GIS based bioenergy study.

Issues needing attention while using GIS for  bioenergy planning

Some  issues that could impact the quality of the GIS outputs are discussed below:

(i)  Pre  and post -GIS  analysis field   visits to  random sample areas are   important to  ascertain the accuracy of  the GIS mapping. Systemic accuracy assessment  using standard methods such as  computation of  error matrix (also called as  confusion matrix)  can   increase the  preciseness of  the GIS output. Bioenergy planning which involves land use  land cover mapping, the minimum mapping accuracy should be 85% (Foody, 2002).
(ii)  The usefulness of the GIS output is influenced by the quality of satellite/digital image used for analysis. For example, good quality land use   land cover map can   be  generated  using medium  resolution image (e.g.  LISS-III satellite  image  of 23.5  meter  spatial  resolution).  However,   to  map  cost- efficient road network for biomass transportation, high res- olution image is necessary (e.g.  LISS-IV image of 5.8 meter spatial resolution) to  extract all  the details of  the major and minor roads of a study area.
(iii)  Precise image processing in  terms of georeferencing, radio- metric  calibration,  noise  removal,  image  enhancement, post-classification smoothening are  other necessary require- ments to  achieve higher mapping accuracy.
(iv)  The  choice of image classification method (manual vs. digi- tal)  can  also  influence the quality of the GIS products.

Therefore, careful consideration of the above points prior to ini- tiating the GIS mapping is essential for precise bioenergy planning.