Betsy McDonald Intro to GIS Final Project Paper December 15, 2013 Introduction With supportive regulatory policies and numerous financial incentives in place, properly sited solar PV development in Massachusetts is becoming more and more attractive. Communities may be looking for opportunities to develop large scale municipal projects, either to support the electric load of municipal buildings, or potentially to site community owned solar projects for residents and other electric customers who otherwise may not be able to site a system on their own property, for technical, monetary, or any other purposes. With this in mind, my analysis centers around the potential to site solar PV on landfills in the Commonwealth. Siting solar PV at landfills can be advantageous, as landfills typically have large, open space, minimal shading, and offer a new use for otherwise unusable or undesirable land. However, there are several factors that can make landfill development challenging, including “permitting restrictions, settlement issues, cap restrictions, system design, distance to interconnection, topography and slope.”1 In 2011, the Massachusetts Department of Energy Resources issued guidance pertaining to municipal development of solid waste landfills in Massachusetts, followed by separate federal level collaborative guidance by the EPA and the National Renewable Energy Laboratory in 2013. Both documents outline opportunities, challenges, and best practices for siting PV on municipal solid waste landfills. With the understanding that a full feasibility analysis would require detailed structural engineering and utility level analysis and plan design that I am unable to complete for this project, my analysis at a basic level attempts to assess the suitability of landfill landscapes for solar PV, by first identifying and then prioritizing the sites with the highest potential (keeping in mind that the analysis is not able to address certain structural and electrical design issues). 1 The Guide to Developing Solar Photovoltaics at Massachusetts Landfills (Rep.). (2011). Retrieved November 5, 2013, from Massachusetts Department of Energy Resources website: http://www.mass.gov/eea/docs/doer/green-communities/pubsreports/pvlandfillguide.pdf Separate analyses were conducted for projects that are already capped, and projects that have yet to be capped. This separation addresses shorter term development plans as well as long term development plans. Landfills that are already closed and capped will face certain structural integration challenges with the existing landfill cap, but if those challenges can be overcome, they could benefit from solar development much sooner when compared to landfills that are still active. Landfills that are still active and not yet capped can benefit from advanced planning for future PV development – if sites with high solar potential are identified before they are closed/capped, landfill owners can make adjustments in their closure plans that will help facilitate solar development and pre-empt certain structural and permitting issues in the future. While I ultimately also compare overall scores, I identified priority rankings by cap status so that state and municipal level resources can be targeted most effectively for comprehensive energy planning needs (short and long term power generation). Spatial Questions There are several spatial questions that pertain to siting solar PV on landfills. The following spatial questions were used to create a site prioritization list for landfills that are already capped and may be most suitable for near-term solar PV development, and a site prioritization list for landfills that are still active but may be most suitable for future/long-term solar PV development (and could benefit from pre-emptively modifying their landfill closure plans). 1. Which landfills currently have large scale solar PV developments onsite, and which landfills already have approved permits for future PV development? (I will want to remove these from consideration, as I want to look at landfills that aren’t yet slated for development). 2. For existing solid waste landfills (active and inactive/capped) that do not yet have solar onsite, how close are they to utility transmission lines? 3. How close are existing solid waste landfills (active and inactive/capped) to areas designated as wetlands? 4. How much solar resource is available at existing landfill facilities? How much shading exists (using canopy cover) onsite? 5. What is the acreage of the landfill, and what is the average slope for that area? 6. In which utility territories are high-potential sites located? 7. Which potential sites are overall most suitable for solar development? (These will be prioritized) Which are least suitable? Relevant Literature Arnette, A. N., & Zobel, C. W. (2011). Spatial analysis of renewable energy potential in the greater southern Appalachian mountains. Renewable Energy, 36, 2785-2798. doi: 10.1016/j.renene.2011.04.024 This article identifies relevant criteria for siting solar facilities, and discusses the use of solar insolation data from NREL, elevation data from USGS, the importance of slope, and land use. The article addresses criticisms of large scale solar PV development, and notes that people may oppose it if it takes up land that could potentially be used for other competing purposes (agriculture, residential/commercial development, etc). As a result of this consideration, the analysis presented in this paper limited the suitable sites to only those identified by NLCD with land covers classified as “barren” (the article considers the “barren” classification to be otherwise unusable for other economic or development purposes). My analysis anticipates a similar criticism but seeks to pre-empt it by only considering landfills and not other spaces or other existing land uses (this isn’t to say that other spaces wouldn’t be appropriate, but rather that I just wanted to prioritize landfills for this assessment, as landfills are already relatively undesirable land uses that could be repurposed through PV development). This article also considered land identified in the Protected Areas Database for the U.S. The database provides information on whether land is restricted for conservation purposes. Guide to Developing Solar Photovoltaics at Massachusetts Landfills (Rep.). (2011). Retrieved November 5, 2013, from Massachusetts Department of Energy Resources website: http://www.mass.gov/eea/docs/doer/green-communities/pubs-reports/pvlandfillguide.pdf This guidebook was extremely valuable in creating the criteria for my suitability assessment, as it provides details regarding PV design and considerations and feasibility characteristics. While it details numerous criteria, I found their inclusion of wetland consideration very helpful, and based on the information provided in this guidebook, I decided to include proximity to wetlands in my criteria. The guidebook indicates that facilities located within 100 feet of wetlands are subject to additional scrutiny, so while I did not exclude these locations from consideration in my analysis, I did give landfills located close to wetlands lower scores. Further, while a suitability analysis is critical for siting landfills, this guidebook indicates numerous other factors for consideration (such as ownership structure) that would factor into a municipalities decision making process. A suitability analysis (like the one I’m demonstrating at the state level) would be only the first step (among many) in the process. Janke, J. (2010). Multicriteria GIS modeling of wind and solar farms in Colorado. Renewable Energy, 35, 2228-2234. doi: 10.1016/j.renene.2010.03.014 This article does not address landfills in particular, but it does detail a process for assessing large scale wind and solar suitability more generally in Colorado. The article outlines the criteria used for site assessment: renewable potential (using NREL annual insolation data), land cover, population density, distance to roads and transmission lines, and cities. The analysis assigned weights based on importance of the criteria, placing the highest value on solar potential (based on the annual insolation) and proximity to transmission lines. Using the reasoning in this article, I gave these two characteristics the most weight in my own analysis. Kiatreungwattana, K., Mosey, G., Jones-Johnson, S., Dufficy, C., Bourg, J., Conroy, A., ... Brown, K. (2013, February). Best practices for siting solar photovoltaics on municipal solid waste landfills (Rep.). Retrieved November 20, 2013, from Environmental Protection Agency website. This document does not specifically address application of GIS, but it does establish a comprehensive list of criteria that are considered best practices for siting solar PV on landfills. Among the relevant criteria, the article recommends a minimum area of two acres of onsite development capacity, a grade or slope of 2-3% (with an upper maximum of 10%), and a half mile distance to the nearest transmission line (this is the ideal, but the article states that a distance of one to three miles may be acceptable). I plan to use the guidelines established by this document as primary determinants for establishing site prioritization. Van Hoesen, J., & Letendre, S. (2010). Evaluating potential renewable energy resources in Poultney, Vermont: A GIS-based approach to supporting rural community energy planning. Renewable Energy, 35(9), 2114-2122. Retrieved from http://dx.doi.org.ezproxy.library.tufts.edu/10.1016/j.renene.2010.01.0. This article looked at development potential for wind, biomass, and solar technologies in a rural town in Vermont. The article is particularly helpful in the detailed methodology – it describes energy potential maps which were created and subsequently used to solicit community input for project development. It also lists the data sources for assessing solar potential including USGS National Elevation Dataset and the National Biomass Carbon Dataset 2000 (used to assess canopy cover). The methodology also incorporates use of state orthophotos. While I used NLCD data for canopy cover and not the National Biomass Carbon Dataset, this article provided a good reference point for developing a systematic assessment of solar potential using GIS. Data Sources Data Set MassDEP Land Disposal of Solid Waste Details Source Year vector 2011 MassGIS http://www.mass.gov/anf/research-and-tech/itserv-and-support/application-serv/office-ofgeographic-informationmassgis/datalayers/towns.html vector 2009 TRANSLINES_ARC MassGIS http://www.mass.gov/anf/research-and-tech/itserv-and-support/application-serv/office-ofgeographic-informationmassgis/datalayers/trnslns.html vector 2007 Massachusetts_DEM_1_ arc_second_NHD_project ed National Map (USGS Data Portal) http://www.mass.gov/anf/research-and-tech/itserv-and-support/application-serv/office-ofgeographic-informationmassgis/datalayers/imgelev5k.html raster Mass GIS http://www.mass.gov/anf/research-and-tech/itserv-and-support/application-serv/office-ofgeographic-informationmassgis/datalayers/depwetlands112000.html vector 2009 Mass DEP http://www.mass.gov/eea/agencies/massdep/se rvice/energy/landfills/renewable-energyprojects-at-closed-landfills.html table 2012 Mass DEP http://www.mass.gov/eea/agencies/massdep/se rvice/energy/landfills/landfills-with-post-closureuse-permits-for-renewables.html table 2013 TOWNS_POLY MassGIS Power Lines DEP Wetlands (1:12,000) WETLANDSDEP_POLY MA Landfill Profiles lfenergy Landfills with PostClosure Use Permits for Renewable Energy Format http://www.mass.gov/anf/research-and-tech/itserv-and-support/application-serv/office-ofgeographic-informationmassgis/datalayers/sw.html SW_LD_POLY Community Boundaries Mass DEM URL Canopy Cover canopy13_011007 NLCD Zone 13 http://www.mrlc.gov/finddata.php raster NREL Insolation/S olar Resource Data DNI NREL http://www.nrel.gov/gis/data_solar.html vector 2001 2005-2009 Steps for data creation, processing, and analysis 1. While many of the data layers were already clipped to Massachusetts, I pre-processed certain layers in order to make sure they were appropriate for my project area (state of Massachusetts). This applied to the NREL insolation data and canopy cover data, both of which needed to be projected to NAD_1983_StatePlane_Massachusetts_Mainland_FIPS_2001. The other MassGIS data layers were already in this projection. 2. I started by selecting by attribute to remove unnecessary fields from my transmission line data set in order to make sure that I am only looking at proximity to power lines and sub stations (and not the ski lifts, landing strips, and airports, etc. that are in the original dataset). 3. Next, two different tables of data (one with info on which landfills already have PV onsite or who have been approved for PV, and one with all landfills in MA with details on additional post-closure uses and permits) were procured from the DEPs website. I had to reformat them, but was ultimately able to merge them into one table using the landfill address as a unique identifier. Once they were merged into one table, I was able to bring that excel worksheet into ArcMap. Using the LNK field as an identifier, I joined my excel table with the Landfill Poly site data set I have from MassGIS. 4. Once all of the DEP data was joined to the solid waste landfill data, I refined this data set to meet some additional basic criteria. I only wanted to consider data for municipal solid waste landfills (the MassGIS data layer had other types of locations, including Ash, C&D Waste, Sludge, Tires, Woodwaste). I selected by attribute so that only MSW landfills remained. I selected by attribute and removed all the sites that already have PV onsite, as I don’t want to consider/assess these for future development (I saved these sites in a separate layer file for future comparison purposes). I also removed landfills that had been capped and already had other uses (ie some have been slated for soccer fields, parking garages). I also selected by attribute to remove all landfills that are under 2 acres in area (as this is the minimum area recommended for site development). 5. I used the Select by Location feature to remove landfills that are more than 3 miles from transmission lines. As this had been identified as an upper limit, I didn’t want to consider landfills that had lines further away than 3 miles. 6. After I pared down the landfill data, I created two data layers – one that has only closed/capped landfills, and one that has only active (and inactive but uncapped) landfills. I wanted to assess suitability for these two categories separately. 7. Using a process similar to what is detailed in some of the suitability reports as well as the exercise we did in class for siting wind facilities in Beijing, I reclassified canopy cover, insolation data, DEM data (to get to slope – low slope is preferable), proximity to transmission lines, and proximity to wetlands, and then create preference grids for each of these based on the criteria I found in the peer-reviewed articles, the Department of Energy Resources report, and the EPA report. Before doing this, I converted all the files to raster grids. See below for criteria ranking details, and see reclassified criteria maps at the end of this document. 8. After creating reclassified feature maps, I did an overlay of the reclassified data sets in order to create final suitability zones. See below for weighting details. I wanted to weight transmission lines the most, with insolation data weighted second highest. I then used the zonal statistics as a table tool to get a mean score for the land area in each landfill polygon. After the zonal statistics tool created tables, I joined them to the appropriate data layers (capped landfills, uncapped landfills, and landfills with existing solar PV onsite). I then sorted each data layer to find the 5 sites with the highest ranking in each of those categories. Due to concerns about the accuracy of canopy data, I did my best to visually locate these landfills using Google Earth and assess shading, within reason. Criteria for Ranking With the aforementioned questions, literature review, and process in mind, I selected the following criteria for determination of suitability: proximity to transmission lines (closer is better, with a max of 3 miles), total solar resource (higher insolation values are better), canopy cover (less cover is better), slope (smaller slope is better), and proximity to wetlands (further from wetlands is better). Overall Criteria Weight 0.30 Criteria for Suitability Proximity to Transmission Lines (miles)* Value 0-.50 .50-.75 .75-1.00 1.00-2.00 2.00-3.00 Score 5 4 3 2 1 Insolation (kWh/m2/day annual) 4.00-4.041 3.80-4.00 3.60-3.80 3.49-3.60 0.0-3.49 5 4 3 2 1 0.25 Canopy Cover (%) 0.00-20.00 20.00-40.00 40.00-60.00 60.00-80.00 5 4 3 2 0.175 80.00-100.00 1 Slope 0.00-2.00 2.00-4.00 4.00-6.00 6.00-8.00 8.00-Max 5 4 3 2 1 0.175 Proximity to Wetlands (feet)* 0.00-100.0 100.00-1,500.00 1,500.00-15,000.00 15,000.00-150,000.00 Max 1 2 3 4 5 0.10 *Input to GIS as Meters Difficulties One of the first difficulties I faced was in preparing the more detailed landfill data (obtained from excel/table data obtained on the DEP’s website). Some of the landfill names and addresses were spelled slightly differently, and this made manual clean-up of the (almost 500) original landfill sites an important first step. Availability of time appropriate data was another challenge in how I was able to conduct the analysis, and a challenge in the ultimate validity of the findings. While my review of literature confirmed that shading was a very important factor, I knew that the canopy data was last updated in 2001, and I struggled deciding whether to use canopy cover at all as part of the analysis. I was unable to determine whether it was not updated in 2006 (with the rest of the NLCD data) due to lack of funding, or whether the data had been determined to be unhelpful. Knowing that shading is a relatively significant factor in determining suitability for solar, I decided to keep the data in my analysis, but weight it less prominently than I would have had I been confident the data was more accurate. As another example of a challenge with evolving data, the data containing post-closure use was last updated in 2012. It is likely that additional landfills have received or applied for post-closure uses (PV or otherwise) since then, and had I had those more recent permit status notes in my updated file, the prioritizations may have turned out differently. Additionally, while the specific criteria were all identified through a literature review, the specific value categories and weights were assigned based on more general understandings that certain criteria (proximity to transmission lines, insolation data, canopy) were more significant than others. I struggled in assigning specific weights, knowing that small differences could have a potentially large difference on my suitability zones and the final list of sites. I also had a hard time deciding whether or not (and if so, how) I would consider proximity to populations in my analysis. I knew that sites ideally wouldn’t be in residential settings, but it also wouldn’t be advantageous to have them too far away. What I couldn’t find was a general rationale for what ideal distance is appropriate. I tried importing census block data several times and played around with different classifications (with very close and very far being ranked with the lowest values), but ultimately I was not confident that I was displaying it appropriately. I also struggled with this because municipal solar projects intended to offset the load of municipal buildings may prefer different distances to community owned solar projects within a given municipality. I think breaking out municipal level goals (instead of lumping these two together) would have helped to a certain degree. I also don’t have a real grasp on how politically viable these projects are and what type of pushback they have received. I ultimately decided to abandon this sixth criteria, acknowledging that my analysis is less accurate as a result. With these difficulties and limitations in mind, it was important for me to remember that these rankings would never be static, and that at best they are a jumping off point (with many future steps to follow) for considering solar development. Overall, I think the analysis demonstrates the basic process of site suitability assessment and the prioritization of geographically ideal sites, keeping in mind that there are numerous additional technical and social factors that may influence site determination, and that the criteria considered in this analysis provide only a basic and approximate understanding of suitability. I do think it could be improved going forward with a meaningful consideration of proximity to high population density locations. Concluding thoughts With the limitations identified above, this analysis presents an approximate prioritization of 5 capped and 5 uncapped MSW landfills that are particularly geographically suited for solar PV development, based on the selected criteria and available data sets. Capped and uncapped landfills were identified separately due to the unique challenges and opportunities each category faces for solar PV development. While no sites achieved the highest score feasible (a mean score of 5, which would have required a perfect score for each individual criteria factor), several of them ranked with total scores over 4. Only 2 (out of 42) sites that currently have solar or are in the process of installing solar achieved a score over 4. Further, several of these landfill locations were not even within the suitability zone identified by this analysis. A comparison of geographically ideal sites for future development with sites that already have or are currently installing solar reinforces the understanding that additional factors influence site selection. Future iterations of this type of analysis may attempt to visualize some of these additional factors. As I mentioned above, going forward, it may be helpful to more explicitly break out municipal goals in installing solar PV (to cover municipal load vs. to help develop participatory community solar projects). More specific delineation of goals may help with considerations of political will or viability and the importance of proximity to urban centers, as examples. Additional Images and Screenshots Individual Factor/Criteria Maps, Reclassified Transmission Lines – Reclass (Data Source: MassGIS) Insolation – Reclass (Data Source: MassGIS, NREL) Slope – Reclass (Data Source: MassGIS, USGS National MAP) Canopy Cover – Reclass (Data Source: MassGIS, NLCD) Wetlands – Reclass (Data Source: MassGIS)