• Ingen resultater fundet

California’s Flexible Ramping Product Need for fast ramping

In document measures for system integration of (Sider 63-66)

California’s high penetration of solar energy along with the high demand peak in the late afternoon has created a challenge known as the “duck curve” in which solar production begins to taper exactly as demand is on the rise (see section 1.5). This has resulted in assets having to ramp at a faster rate and over greater ranges of output. Current38 solar and wind penetration levels have only necessitated about 10 to 12 GW of flexible procurement in CAISO. However, with California’s rapidly scaling renewable portfolio standard which reaches 50% in 2030, flexible resource adequacy requirements would rise by roughly 17 GW under a scenario of equal growth in wind and solar39.

Figure 16: CAISO Monthly Flexible Demand Scenarios vs. Current Flexible Supply. Horizontal axis depicts months from January to December.

Dispatchers have begun scheduling differently to accommodate, because often the units scheduled under day-ahead markets (DAM) lack sufficient ramping capability and flexibility to handle the afternoon ramp up, and the capacity available to be dispatched in real-time is also insufficient to cover fluctuations in solar/wind between different 5-minute dispatch periods.

Flexible Ramping product

California created their flexible ramping product (FRP), which allowed CAISO to co-optimize

38 Reference based on year 2017

39 https://www.greentechmedia.com/articles/read/can-california-conquer-the-next-phase-of-renewables-integration#gs.FpkaZKw

64

their flexible capacity scheduling with their energy market. First, CAISO implemented a flexible ramping sufficiency test to identify how much flexibility was required, considering things like renewable and demand forecast errors and contingencies that resulted in generators previously scheduled being unable to ramp. After several years trial operation using an administrative pricing mechanism, on November 1st 2016, CAISO implemented two market products in the 15- and 5-minute markets: Flexible Ramp Up and Flexible Ramp Down. In addition, CAISO extended the existing flexible ramping sufficiency test to the entire EIM footprint, to ensure feasible ramping capacity for all real-time interchange schedules.

The result of this increased flexibility has been less out-of-merit order dispatching (lower cost), minimizing the use of reserve margins for planned excursions (greater reliability), and less solar curtailment (less emissions). While the flexible ramping product may be used to attract new flexible resources when the system is approaching a shortfall, given today’s sufficiency levels, these flexible incentives also help prevent retirement of existing critical flexible capacity in the face of solar-induced revenue erosion in wholesale markets40. Plants that are typically scheduled for these flexibility ramps are often plants already on the margin and have been struggling to meet their going forward costs. Due to the new incentive

provided by the flexible ramping product, they have further refined their operation to be as flexible as possible, even two-shifting operation (starting up and shutting down twice in a day).

Total net payments to generators in the ISO and energy imbalance market areas for providing flexible ramping capacity in 2017 were about $25 million. As market players better understood the products and markets became more competitive, monthly total payments decreased during the year from around $3 million per month between January and May to around $1.4 million between June and December. Although flexible ramping payments increased with the implementation of the flexible ramping products, payments per megawatt-hour of load remained low, with average net payments per megawatt-hour of load were about $0.07/MWh during 2017 (figure 17)41. For comparison, payments for ancillary services in the ISO were about $0.75/MWh of load during the same time period.

Whether FRP reduced the overall AS-energy market cost is hard to quantify due to the difficulty of controlling for all the variables influencing electricity prices, FRP implementation issues, and an increase in operating reserve requirement in 2017. Ancillary service costs increased to $0.75/MWh of load served in 2017 from $0.52/MWh in 2016. This represents an increase from about 1.6% of total wholesale energy costs in 2016 to 1.9% in 2017, which are the highest yearly values since 2011, both as a percentage of wholesale energy costs and per megawatt-hour of load42. While the increase in reserve requirement drove up the volume and price for procuring regulation, the need for regulation seemed to decrease with the implementation of FRP in 2017 (figure 18), however this result suffers from a calculation error which resulted in an underestimate for the need of upward flexible ramping capacity during some ramping intervals43. The error was corrected in February 2018 and these future results will be necessary to verify FRP’s effectiveness.

40 https://www.greentechmedia.com/articles/read/can-california-conquer-the-next-phase-of-renewables-integration#gs.FpkaZKw

41 http://www.caiso.com/Documents/2017AnnualReportonMarketIssuesandPerformance.pdf 42 Ibid.

43 http://www.caiso.com/Documents/2017AnnualReportonMarketIssuesandPerformance.pdf

Part I USA

65

Figure 17: Monthly flexible ramping payments by balancing area

Figure 18: Monthly average day-ahead regulation require.

Beyond FRP’s integration into real-time markets, CAISO also incorporated flexibility needs assessments into its resource planning process. It requires utilities to procure sufficient flexible resources for future delivery years and ensures flexible resources meet technical criteria for ramp speed and response time and are required to be available and bid into the CAISO energy markets44.

44 https://www.greentechmedia.com/articles/read/can-california-conquer-the-next-phase-of-renewables-integration#gs.zHycEoc

47 Figure 3-7 Monthly average day-ahead regulation require.

66 https://www.nrel.gov/docs/legosti/old/7803.pdf 67 https://www.nrel.gov/docs/fy12osti/50907.pdf

68 http://greeningthegrid.org/trainings-1/presentation-an-introduction-to-wind-and-solar-power-forecasting-1 RE generation forecast improvement to transit flexible ramping need

As FRP is specifically designed to deal with the flexibility incurred due to forecast errors, increasing renewable forecast accuracy will directly reduce the need of FRP, and more broadly reduce the need for other flexibility.

The benefit of wind forecast improvements are well proven by variousUSAresearchers.6667Generally, more accurate forecasts can:68

Reduce reserve levels: Regulation reserve; Flexible/load following reserve

Improve unit commitment and dispatch efficiency: Utilize the least expensive units; Less “mileage” on operating units; Less starting of gas turbines and other fast acting units,

ReduceVREcurtailment The case of Xcel

46 Figure 4-6 Monthly flexible ramping payments by balancing area

Whether FRP reduced the overall AS-energy market cost is hard to quantify due to the difficulty of controlling for all the variables influencing electricity prices, FRP implementation issues, and an increase in operating reserve requirement in 2017. Ancillary service costs increased to $0.75/MWh of load served in 2017 from $0.52/MWh in 2016. This represents an increase from about 1.6% of total wholesale energy costs in 2016 to 1.9% in 2017, which are the highest yearly values since 2011, both as a percentage of wholesale energy costs and per megawatt-hour of load.63 While the increase in reserve requirement drove up the volume and price for procuring regulation, the need for regulation seemed to decrease with the implementation of FRP in 2017 (Figure 4.7), however this result suffers from a calculation error which resulted in an underestimate for the need of upward flexible ramping capacity during some ramping intervals.64 The error was corrected in February 2018 and these future results will be necessary to verify FRP’s effectiveness.

Beyond FRP’s integration into real-time markets, CAISO also incorporated flexibility needs assessments into its resource planning process. It requires utilities to procure sufficient flexible resources for future delivery years and ensures flexible resources meet technical criteria for ramp speed and response time and are required to be available and bid into the CAISO energy markets.65

63 ibid

64 http://www.caiso.com/Documents/2017AnnualReportonMarketIssuesandPerformance.pdf

65 https://www.greentechmedia.com/articles/read/can-california-conquer-the-next-phase-of-renewables-integration#gs.zHycEoc

66

RE generation forecast improvement to

In document measures for system integration of (Sider 63-66)