II. Background
1. The Electricity Grid
Due to the contemporary structure of electric grids, electricity must be generated at the time of use, which causes inefficiencies that exacerbate the associated impact on anthropogenic climate change. Even everyday conditions such as diurnal fluctuations in electricity demand are a source of avoidable emissions (Dell and Rand 2001). The inclusion of ES systems on a grid has the potential to decouple electricity supply from demand thereby reducing the impact of such inefficiencies (Chen et al 2009, Dell and Rand 2001).
Chen et al (2009) indicate diurnal and annual demand fluctuations not only cause generation inefficiencies but also require that generation capacity be over-built to meet peak demand that may only last a few hours each year. With sufficient ES capacity to meet such peak-demand the construction of additional primary generation capacity can be delayed or avoided (Dell and Rand 2001).
Dell and Rand (2001) suggest that peak-shaving and load-leveling with ES can reduce the need to maintain plants in spinning reserve (i.e. generating electricity at sub-nominal values in case the plant needs to be called into regular service with short notice due to an unplanned plant failure or an unexpected increase in demand) to avoid a short-term shutdown. When fossil fuel plants are operating in spinning reserve, the GHG emissions per kWh of generation is greater than emissions during optimal generation (Voorspools and D’haeseleer 2000). Furthermore; this effect is not insignificant. As one report from the PJM Interconnection (a regulatory authority in the USA) shows annually the average GHG emissions from so-called ‘Marginal Units’ are over 50% greater than the overall average emissions of all generators (PJM 2009). This leaves a significant amount of room for emissions reductions. By charging ES systems during low-demand and allowing ES systems to meet demand during peak conditions or to be held in reserve for unplanned plant outages, plants can be maintained at optimal generation levels by running at a constant or near-constant rate (load-leveling or peak-shaving respectively) (Chen et al 2009).
Spinning reserve is one of two types of ancillary services required to ensure that the electricity grid seamlessly meets consumer demand. The second type of ancillary service is commonly called regulation (or frequency regulation or regulation power). Kempton et al (2008) explain that regulation maintains the grid at its optimal frequency (i.e. 60 Hz in the USA and 50 Hz in the rest of the world). When electricity generation exceeds demand the frequency increases. Contrapositively, the frequency decreases when electricity demand exceeds generation. In order to maintain the grid at the optimal frequency, ancillary service providers are called upon to provide regulation up (i.e. increasing generation or decreasing demand) and regulation down (i.e. decreasing generation or increasing demand) services (Kempton et al 2008). Currently, traditional generation plants are tasked with responding to regulation demands; however, as Shelton (2010) illustrates traditional plants are typically not capable of responding quickly to regulation requirements. Additionally, traditional plants acting in regulation mode are subject to the same inefficiencies as described in the previous paragraph. Thus, regulation services provide another opportunity for emissions reductions.
In addition to the limitations of traditional energy production, ES appears likely to play an even greater role with regard to future development plans and attempts to mitigate climate change. Perhaps most significantly, as developed nations look to integrate so-called ‘emission-free’ renewable energy technologies with intermittent generation into their energy portfolio (e.g. NCSC 2009), the integration of ES systems on the utility grid may be not only desirable but necessary for the sake of practicality and the economic viability of a large-scale implementation of renewable generating capacity (e.g. Pickard et al 2009, Aguado et al 2009, Benitez et al 2008).
Finally, Bayod-Rújula (2009) and Verhaegh et al (2010) suggest that the future of the electricity grid in developed countries will likely involve increased distributed (non-centralized) generation and/or the wide-scale use of EVs and residential heat-pumps, which may vastly alter the nature of the contemporary diurnal supply and demand cycles. In the US in particular mass production of EVs seems imminent within the next several years (e.g. Woody and Krauss 2010). Thus, earlier studies that have not consider these developments will need to be reexamined or taken with caution.
a. Generation Technologies
Voorspools et al (2000) have suggested that studies analyzing GHG emissions associated with ‘emission-free’ technologies need to take into account indirect emissions embedded in construction in addition to the direct emissions from the fuel cycle, which is traditionally the limit of the scope of energy generation analyses. Weisser (2007) explains that all energy systems in use today have some associated GHG emissions, irrespective of the generation technology used. Performing a full life cycle analysis (LCA) of various generation technologies allows for a more comprehensive understanding of the actual GHG emissions associated with the electricity being transmitted through the grid. A comprehensive LCA tabulates the cumulative emissions (i.e. those associated with the entire life cycle of the generator), which is the sum of direct (i.e. emissions created during the generation process) and indirect (i.e. emissions created ‘upstream’ of the generation process such as fuel processing and plant construction and emissions created ‘downstream’ of the generation process such as plant decommissioning and fuel disposal) emissions (Weisser 2007). Further, even if the same generation technology is used, the level of associated GHG emissions can vary greatly depending on the source of the fuel (Jaramillo et al 2007).
For the purposes of this study, the range of generation technologies was limited to those in regular usage in the PJM territory and those that are expected to be added within the next two decades. The combination of nuclear, coal, natural gas, oil, hydroelectric power currently accounts for more than 95% of the electricity generated in the PJM territory (see section III.2.). Meanwhile, solar and especially wind power show significant promise for future expansion (Boston and Mansoor 2010). As such, finding reliable LCA GHG emissions values for all of these generation technologies is crucial for assessing the impact of installing ES on the electricity grid.
b. Consumption Patterns
Even from the earliest development of the electric utility grid in the USA starting in the 1890’s, generation plants have been constructed and operated with the expectation that demand will be met by altering supply in a real-time manner (Schainker 2010). This means that for any alterations to the utility grid to be successfully implemented the needs of society must still be met. Thus, for the most part as utility companies and regulators plan for future developments, traditional consumption patterns can be assumed.
In PJM territory, which is not uncommon for developed countries, consumption patterns tend to follow a number of general patterns. Diurnally, demand tends alternate between on-peak (when demand is high) and off-peak (when demand is relatively low) between the day and night respectively. On a weekly basis, demand tends to be higher during the work week (i.e. Monday through Friday) and lower during weekends and on holidays (PJM 2009). Furthermore, most grid systems tend to experience seasonal fluctuations in demand annually; however, these fluctuations are harder to generalize across different territories as they depend on variables such as the local climate, local industries, and the preferred choice of air conditioning technologies in the territory (e.g. Lund and Münster 2003). With these considerations in mind, this study accounts for diurnal and weekly consumption patterns but does not attempt to quantify seasonal differences.
Previous Post: Introduction
No comments:
Post a Comment