Rob J Hyndman

Rob J Hyndman and Shu Fan

IEEE Trans­ac­tions on Power Sys­tems, 2010, 25(2), 1142–1153

Abstract: Long-​​term elec­tri­city demand fore­cast­ing plays an import­ant role in plan­ning for future gen­er­a­tion facil­it­ies and trans­mis­sion aug­ment­a­tion. In a long term con­text, plan­ners must adopt a prob­ab­il­istic view of poten­tial peak demand levels, there­fore dens­ity fore­casts (provid­ing estim­ates of the full prob­ab­il­ity dis­tri­bu­tions of the pos­sible future val­ues of the demand) are more help­ful than point fore­casts, and are neces­sary for util­it­ies to eval­u­ate and hedge the fin­an­cial risk accrued by demand vari­ab­il­ity and fore­cast­ing uncer­tainty. This paper pro­poses a new meth­od­o­logy to fore­cast the dens­ity of long-​​term peak elec­tri­city demand.

Peak elec­tri­city demand in a given sea­son is sub­ject to a range of uncer­tain­ties, includ­ing under­ly­ing pop­u­la­tion growth, chan­ging tech­no­logy, eco­nomic con­di­tions, pre­vail­ing weather con­di­tions (and the tim­ing of those con­di­tions), as well as the gen­eral ran­dom­ness inher­ent in indi­vidual usage. It is also sub­ject to some known cal­en­dar effects due to the time of day, day of week, time of year, and pub­lic holidays.

We describe a com­pre­hens­ive fore­cast­ing solu­tion in this paper. First, we use semi-​​parametric addit­ive mod­els to estim­ate the rela­tion­ships between demand and the driver vari­ables, includ­ing tem­per­at­ures, cal­en­dar effects and some demo­graphic and eco­nomic vari­ables. Then we fore­cast the demand dis­tri­bu­tions using a mix­ture of tem­per­at­ure sim­u­la­tion, assumed future eco­nomic scen­arios, and resid­ual boot­strap­ping. The tem­per­at­ure sim­u­la­tion is imple­men­ted through a new sea­sonal boot­strap­ping method with vari­able blocks.

The pro­posed meth­od­o­logy has been used to fore­cast the prob­ab­il­ity dis­tri­bu­tion of annual and weekly peak elec­tri­city demand for South Aus­tralia since 2007. We eval­u­ate the per­form­ance of the meth­od­o­logy by com­par­ing the fore­cast res­ults with the actual demand of the sum­mer ²⁰⁰⁷⁄₀₈.

Keywords: Long-​​term demand fore­cast­ing, dens­ity fore­cast, time series, simulation.

Online art­icle