Tuesday, December 7, 2010

Dissertation: Methodology (1 of 8) - Data Collection


 III.    Methodology

1.   Data Collection

From the outset of this study, which sought to analyze and compare a wide variety of technologies at various stages of research, development, deployment, and use, it was clear that ‘data collection’ would be necessarily complicated. Initially, the expectation was that sufficient data could be collected directly from owners, operators, and manufacturers of ES technologies; however, it soon became apparent that additional supplementary data sources would be required to compile a comprehensive set of parameters for modeling purposes. Thus, a manifold process of data collection was put into action. This process included: (a) identifying and contacting owners, operators, and manufacturers of ES technologies as well as electric utility operators in the model region (i.e. the PJM Interconnection territory in the USA), (b) developing and distributing a suitable survey, and (c) identifying and interpreting supplementary data sources.

a.    Direct Sources

Since the majority of data required to develop modeling parameters is based on technical data related to the performance of ES technologies (many of which are still in the developmental or even research phases), it was expected that it would be difficult to collect first-hand data. Furthermore, due to the nature of grid operations, the time scales needed for analysis, and the limited timeline on which this study needed to be performed, it was deemed necessary to rely upon second-hand information regarding the technical specifications of ES technologies and the electricity grid rather than direct measurements.

Thus, to most directly access the necessary data, it was decided that surveying those who used and manufactured the equipment in question was the most prudent approach. In order to identify appropriate survey recipients, a contact list was compiled by: (1) searching for relevant terms on the Google internet search engine, (2) noting companies that were clearly identified within academic journal articles (e.g. those listed in the Reference section), (3) identifying key individuals who presented at the 2010 PJM-EPRI Energy Storage Summit, and (4) noting the utility companies and regulators listed on the Pennsylvania Public Utility Commission Website (i.e. http://www.puc.state.pa.us/electric/electric_companies.aspx).

b.   Survey Development, Distribution, & Responses

The survey (see Appendix A) was developed with energy storage professionals and utility operators in mind. As such, it includes a brief introduction and some key instructions to explain the purpose of the survey, describes how to complete and submit the survey, and encourages survey recipients to participate. As a limited incentive, all survey participants were offered a free copy of the final version of this dissertation. However, since the intended participants were those who were already knowledgeable of ES and utility systems, there was no explicit technical information provided to the participants within the survey.

The survey distribution process included multiple stages. The first stage (prior to contacting survey participants) included gathering contact information (see section III.1.a. above) and posting information about the dissertation study on a website (i.e. http://seandiamondsustainability.blogspot.com/). The second stage involved making initial contact with potential participants, which was delivered through a variety of media depending on what contact information was available. The majority of participants were contacted prior to the release of the survey using a letter sent via the postal service. The remainder of the participants was either contacted via email or telephone. The third stage involved distributing the actual survey via email and posting it on the website. After the third stage, follow up emails or phone calls were made to participants to encourage them to participate and answer questions or concerns the participants might have.

On a separate note, it should be addressed that though respondents were not given any direct incentive to provide a particular type of answer (i.e. “better” figures) and they were ensured that there responses would not directly identify them, the potential for respondents to give a best-case-scenario response remained a possibility. This was especially the case for ES manufacturers who could stand to benefit financially by making their product seem marginally more beneficial than a competing technology or product. Further, even if there was no intent to provide augmented responses, it should be understood that many of the technologies in question have not necessarily been exhaustively field tested through massive commercial deployment. Unfortunately, due to the nature of this study, there was no a priori way to ensure or measure the reliability of the responses provided. Thus, all survey responses have been taken as truthful and reasonably accurate.

Responses to the proactive, multi-stage survey approach varied. The majority of participants who were contacted did not respond at all. However, some (approximately 20%) responded positively to the initial contact and follow up attempts. Unfortunately, the final response rate (i.e. those who completed and submitted a survey response by or soon after the submission deadline) was abysmally low (approximately 6%) and did not yield sufficient data upon which to establish modeling parameters. Since the survey responses were ultimately not useful for the end results of this study, they were not included in attempts develop model parameters.

c.    Supplementary Sources

The initial expectation of using supplemental sources of information to fill in any gaps left by the survey responses ended up being an underestimation of their value. Ultimately, secondary sources (i.e. academic journal articles) and data compiled by the PJM regulatory authority for other purposes (e.g. public relations reports) were indispensible sources of data in this study. Furthermore, despite their label as supplementary sources, the entirety of the modeling parameters were established based on the information obtained from these sources. For a full list of the data compiled from supplementary sources see Appendix B (information regarding ES technologies) and Appendix C (information regarding the model region).

No comments:

Post a Comment