### Abstract

The variability inherent in wind power production will require increased flexibility in the power system, when a significant amount of load is covered with wind power. Standard deviation (σ) of variability in load and net load (load net of wind) has been used when estimating the effect of wind power on the short term reserves of the power system. This method is straightforward and easy to use when data on wind power and load exist. In this paper, the use of standard deviation as a measure of reserve requirement is studied. The confidence level given by ±3-6 times σ is compared to other means of deriving the extra reserve requirements over different operating time scales. Also taking into account the total variability of load and wind generation and only the unpredicted part of the variability of load and wind is compared. Using an exceedence level can provide an alternative approach to confidence level by standard deviation that provides the same level of risk. The results from US indicate that the number of σ that result in 99% exceedence in load following time scale is between 2.3-2.5 and the number of σ for 99.7% exceedence is 3.4. For regulation time scale the number of σ for 99.7 % exceedence is 5.6. The results from the Nordic countries indicate that the number of σ should be increased by 67-100% if better load predictability is taken into account (combining wind variability with load forecast errors).

Original language | English (US) |
---|---|

Pages (from-to) | 355-378 |

Number of pages | 24 |

Journal | Wind Engineering |

Volume | 32 |

Issue number | 4 |

State | Published - 2008 |

### Fingerprint

### ASJC Scopus subject areas

- Energy Engineering and Power Technology
- Renewable Energy, Sustainability and the Environment

### Cite this

*Wind Engineering*,

*32*(4), 355-378.

**Using standard deviation as a measure of increased operational reserve requirement for wind power.** / Holttinen, Hannele; Milligan, Michael; Kirby, Brendan; Acker, Tom L; Neimane, Viktoria; Molinski, Tom.

Research output: Contribution to journal › Article

*Wind Engineering*, vol. 32, no. 4, pp. 355-378.

}

TY - JOUR

T1 - Using standard deviation as a measure of increased operational reserve requirement for wind power

AU - Holttinen, Hannele

AU - Milligan, Michael

AU - Kirby, Brendan

AU - Acker, Tom L

AU - Neimane, Viktoria

AU - Molinski, Tom

PY - 2008

Y1 - 2008

N2 - The variability inherent in wind power production will require increased flexibility in the power system, when a significant amount of load is covered with wind power. Standard deviation (σ) of variability in load and net load (load net of wind) has been used when estimating the effect of wind power on the short term reserves of the power system. This method is straightforward and easy to use when data on wind power and load exist. In this paper, the use of standard deviation as a measure of reserve requirement is studied. The confidence level given by ±3-6 times σ is compared to other means of deriving the extra reserve requirements over different operating time scales. Also taking into account the total variability of load and wind generation and only the unpredicted part of the variability of load and wind is compared. Using an exceedence level can provide an alternative approach to confidence level by standard deviation that provides the same level of risk. The results from US indicate that the number of σ that result in 99% exceedence in load following time scale is between 2.3-2.5 and the number of σ for 99.7% exceedence is 3.4. For regulation time scale the number of σ for 99.7 % exceedence is 5.6. The results from the Nordic countries indicate that the number of σ should be increased by 67-100% if better load predictability is taken into account (combining wind variability with load forecast errors).

AB - The variability inherent in wind power production will require increased flexibility in the power system, when a significant amount of load is covered with wind power. Standard deviation (σ) of variability in load and net load (load net of wind) has been used when estimating the effect of wind power on the short term reserves of the power system. This method is straightforward and easy to use when data on wind power and load exist. In this paper, the use of standard deviation as a measure of reserve requirement is studied. The confidence level given by ±3-6 times σ is compared to other means of deriving the extra reserve requirements over different operating time scales. Also taking into account the total variability of load and wind generation and only the unpredicted part of the variability of load and wind is compared. Using an exceedence level can provide an alternative approach to confidence level by standard deviation that provides the same level of risk. The results from US indicate that the number of σ that result in 99% exceedence in load following time scale is between 2.3-2.5 and the number of σ for 99.7% exceedence is 3.4. For regulation time scale the number of σ for 99.7 % exceedence is 5.6. The results from the Nordic countries indicate that the number of σ should be increased by 67-100% if better load predictability is taken into account (combining wind variability with load forecast errors).

UR - http://www.scopus.com/inward/record.url?scp=67650082529&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67650082529&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:67650082529

VL - 32

SP - 355

EP - 378

JO - Wind Engineering

JF - Wind Engineering

SN - 0309-524X

IS - 4

ER -