2. Model Region and Scenario Selection
Voorspools and D’haeseleer (2000) stress the need for a simulation tool, stating: “Since it would be impractical to constantly monitor the instantaneous composition of the power system and to calculate (or measure) the corresponding emissions… [f]or studies or scenarios carried out for future or hypothetical developments, monitoring is not even an option and, hence, a simulation tool is essential.” However, a simulation is only really useful to the extent that it is analogous to situations in the real world. In order to maximize the similarities between the model developed in this study and the real world, this study based its characteristics around a real utility grid (i.e. the PJM Interconnection in the USA) and developed scenarios based around plausible alterations to the electricity generation and ES capacity within the established timeline (i.e. prior to 2030).
Initially, this study focused on the PJM region due to its relative familiarity to the author. However, upon even superficial inspection the PJM Interconnection turned out to be perhaps the most appropriate choice. As Boston and Mansoor (2010) explain during an introductory presentation at the PJM-EPRI Energy Storage Summit held in April 2010 the PJM is actively partnering with other companies and organizations to pursue increased deployment of ES within its territory in the immediate future. This ES-friendly sentiment is unequivocally expressed by PJM’s CEO with the need for “Storage, Storage, and More Storage” with regard to the future development plans of PJM (Boston and Mansoor 2010). Furthermore, since PJM covers 168,500-square-miles and roughly 51 million residents in the Northeastern and Midwestern USA (PJM 2010), it is likely that any decisions that PJM makes with regard to future grid development will significantly impact the decisions made by (and subsequently the GHG emissions of) other grid operators across the USA if not the world at-large. Thus, by choosing the PJM territory as a model region the results from this study should be more widely, though less precisely, applicable to the real world beyond the geographic limits of the PJM itself.
This paragraph briefly summarizes some of the characteristics of the PJM territory that were immediately relevant in setting up the model and using PJM as the model region (refer to Appendix C for a full list of PJM statistics considered in this study). In 2009 the annual peak load for the PJM was nearly 145 GW and approximately 729,000 GWh of electricity was generated (Boston and Mansoor 2010). Thus, the PJM has an annual load factor (actual electricity generated divided by the amount of energy that would have been generated if electricity demand was constantly at the peak load; see Söderholm 2001 for a further explanation of load factor) of approximately 0.58. Schainker (2010) states that most of the PJM territory has geography that is suitable for ES (specifically CAES) including a number of depleted gas fields. Finally – and importantly for calculating GHG emissions – ‘marginal’ power plants (i.e. those running at suboptimal generation) produce approximately 60% (with a range of approximately 45-78%) more GHG emissions per unit of electricity generation on average than those running at optimal capacity (PJM 2009).
The four scenarios (plus one baseline scenario) evaluated in this study are based upon possible alterations to the generation capacity utilized within the PJM.
The baseline scenario (BL) generalized somewhat but matched the percentage of generation technologies used during 2009 as closely as possible. Traditional non-intermittent generators met over 95% (most of which was nuclear (~35%), coal (~50%), and natural gas(~10%)) of the electricity demand in the PJM during 2009 (see Appendix C for exact figures). The PJM currently has about 2.8 GW of installed wind capacity and much less than 1 GW of installed solar capacity (Boston and Mansoor 2010). It should also be noted that unlike all of the other scenarios, in BL the model assumed that any gap between the electricity demand and supply was filled by importing additional electricity from neighboring grids. Since information was not collected about the neighboring grids, the assumption was made that all additional electricity would have the same GHG emissions rate as the average of the electricity generated within the territory.
Scenario One (S1) used the same generation percentages as BL; however, S1 used ES capacity rather than importing electricity to meet additional demand. Scenario Two (S2), Scenario Three (S3), and Scenario Four (S4) altered the generation capacities used by the grid. S2 and S3 both have an installed wind capacity of nearly 40 GW, which is similar to the amount of wind capacity that is already committed to be installed in PJM in the near future (Boston and Mansoor 2010). In S2 the increased wind capacity is offset by a proportionate decrease in nuclear, coal, and natural gas capacity. Whereas, in S3 nuclear capacity is actually increased and coal and natural gas capacity are significantly decreased. In S4 both wind and solar capacity are increased significantly beyond the levels simulated in S2 and S3 with nuclear, coal, and natural gas capacity decreased below BL.
In BL and all other scenarios the demand curve is kept the same (see section III.3. for more detail). In S2, S3, and S4 oil capacity (which accounts for about 0.25% of BL capacity) was dropped to zero. In contrast in S2, S3, and S4, traditional hydroelectric capacity (which accounts for about 0.1% of BL capacity) was maintained at a constant level. Please see Table 2 for a comparison of all generation capacities relative to BL.
Table 2: This table compares the quantity of generation capacity in scenarios S1, S2, S3, and S4 to the generation capacity of the same generation type in scenario BL.
Generation Type | S1 | S2 | S3 | S4 |
Nuclear | x1.00 | x0.67 | x1.67 | x0.29 |
Coal | x1.00 | x0.66 | x0.12 | x0.25 |
Natural Gas | x1.00 | x0.34 | x0.21 | x0.43 |
Oil | x1.00 | x0.00 | x0.00 | x0.00 |
Hydro | x1.00 | x1.00 | x1.00 | x1.00 |
Wind | x1.00 | x14.29 | x14.29 | x42.86 |
Solar | x1.00 | x1.00 | x1.00 | x150.00 |
Generally speaking, S1 represents the present conditions of the grid if ES were added without any additional changes. S2 roughly represents business-as-usual conditions with the addition of ES during the time period examined in this study. S3 represents a possible ‘nuclear future’, which is favored by some policymakers attempting to reduce GHG emissions, where ES has been included into the mix. S4 represents a possible ES-anchored ‘renewable future’, which follows a more ambitious attempt to reduce GHG emissions than is currently being pursued in the PJM territory. Thus, with the exception of increasing the share of fossil fuel generation capacity (which is already about 60% of overall generation) this study attempts to cover all plausible future development plans within the model region.
It should be noted that even though S1 has a directly analogous baseline scenario (i.e. BL), there is no such equivalent baseline scenario for S2, S3, or S4. This is because if S3 or S4 are to be pursued, many predict that some form ES will be a mandatory addition to the utility grid (e.g. Salgi and Lund 2008, Lund and Salgi 2009, and Barreto et al 2003). Also, as McIntosh (2010) discusses, utility grids with greater percentages of intermittent renewable generation capacity than the model region – such as the California ISO that has a smaller percentage than that represented in S2 – tend to have electric transmission systems that are under significant stress caused by intermittency issues. Further, during the development and debugging of the model, preliminary test trial runs that did not incorporate ES capabilities in situations similar to S2 produced electricity supply curves that did not match demand, which in the real world would not be acceptable to contemporary electricity consumers in developed countries. Thus, it seemed nonsensical (and unreliable) to try to incorporate such situations in the final results of this study. Instead, it should be sufficient to note that once the inclusion of a certain critical percentage of a grid’s generating capacity consists of intermittent renewable generators either ES or some other method of intervention must be made to maintain a stable grid. Otherwise, the traditional electric utility model of ‘on-demand’ electricity supply would need to be abandoned.
Previous Post: Methodology (1 of 8) - Data Collection
Next Post: Methodology (3 of 8) - Modeling
No comments:
Post a Comment