Energy Forecasting Methods - Purdue University

2y ago
26 Views
2 Downloads
236.59 KB
48 Pages
Last View : 18d ago
Last Download : 2m ago
Upload by : Kelvin Chao
Transcription

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateEnergy Forecasting MethodsPresented by:Douglas J. GothamState Utility Forecasting GroupEnergy CenterPurdue UniversityPresented to:Indiana Utility Regulatory CommissionIndiana Office of the Utility Consumer CounselorNovember 15, 2007

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateOutline Modeling techniques Projecting peak demand from energyforecasts Determining capacity needs fromdemand forecasts Incorporating load management andconservation measures Uncertainty

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateUsing the Past to Predict the Future What is the next number in the followingsequences?– 0, 1, 4, 9, 16, 25, 36, 49, .– 0, 1, 3, 6, 10, 15, 21, 28, .– 0, 1, 2, 3, 5, 7, 11, 13, .– 0, 1, 1, 2, 3, 5, 8, 13, . These types of problems are at theheart of what forecasters do

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateA Simple 0960?940?900?1040980920123456

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateA Little More 30012001100?1000?9001?23456

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateMuch More 992017000?16000?

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateMuch More Difficult The numbers on the previous slide werethe summer peak demands for Indianafrom 2000 to 2005. They are affected by a number offactors– Weather– Economic activity– Price– Interruptible customers called upon– Price of competing fuels

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup 02200019981996199419921990198819861984198201980 How do we finda pattern inthese peakdemandnumbers topredict thefuture?25000

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateThe Short Answer

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateMethods of Forecasting Time Series– trend analysis Econometric– structural analysis End Use– engineering analysis

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateTime Series Forecasting Linear Trend– fit the best straight line to the historical data and assume thatthe future will follow that line (works perfectly in the 1stexample)– Many methods exist for finding the best fitting line, the mostcommon is the least squares method. Polynomial Trend– Fit the polynomial curve to the historical data and assumethat the future will follow that line– Can be done to any order of polynomial (square, cube, etc)but higher orders are usually needlessly complex Logarithmic Trend– Fit an exponential curve to the historical data and assumethat the future will follow that line (works perfectly for the 2ndexample)

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateGood News and Bad News The statistical functions in most commercialspreadsheet software packages will calculatemany of these for you These may not work well when there is a lotof variability in the historical data If the time series curve does not perfectly fitthe historical data, there is model error.There is normally model error when trying toforecast a complex system.

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateMethods Used to Account forVariability Modeling seasonality/cyclicality Smoothing techniques– Moving averages– Weighted moving averages– Exponentially weighted moving averages Filtering techniques Box-Jenkins

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateEconometric Forecasting Econometric models attempt to quantify therelationship between the parameter ofinterest (output variable) and a number offactors that affect the output variable. Example– Output variable– Explanatory variable Economic activityWeather (HDD/CDD)Electricity priceNatural gas priceFuel oil price

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateEstimating Relationships Each explanatory variable affects the outputvariable in different ways. The relationshipscan be calculated via any of the methodsused in time series forecasting.– Can be linear, polynomial, logarithmic Relationships are determined simultaneouslyto find overall best fit. Relationships are commonly known assensitivities.

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateEnd Use Forecasting End use forecasting looks at individualdevices, aka end uses (e.g., refrigerators) How many refrigerators are out there? How much electricity does a refrigerator use? How will the number of refrigerators changein the future? How will the amount of use per refrigeratorchange in the future? Repeat for other end uses

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateThe Good News Account for changes in efficiency levels (newrefrigerators tend to be more efficient thanolder ones) both for new uses and forreplacement of old equipment Allow for impact of competing fuels (naturalgas vs. electricity for heating) or forcompeting technologies (electric resistanceheating vs. heat pump) Incorporate and evaluate the impact ofdemand-side management/conservationprograms

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateThe Bad News Tremendously data intensive Primarily limited to forecasting energyusage, unlike other forecasting methods– Most long-term planning electricityforecasting models forecast energy andthen derive peak demand from the energyforecast

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateExample State Utility Forecasting Group (SUFG)has electrical energy models for each of8 utilities in Indiana Utility energy forecasts are built up fromsectoral forecasting models– residential (econometric)– commercial (end use)– industrial (econometric)

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateAnother Example The Energy Information Administration’sNational Energy Modeling System (NEMS)projects energy and fuel prices for 9 censusregions Energy ransportation

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateSUFG Residential Sector Model––––demographicshouseholdshousehold incomeenergy 99200350000Ye a r sAnnual Use per Non-Electric Space HeatingCustom 9871991199519992003 Major forecastdrivers2500020000Year– electric– non-electricAnnual Use per Electric Space HeatingCustom erYear Residential sectorsplit according tospace heatingsourceYe a r s

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateResidential Model SensitivitiesSource: SUFG 2005 Forecast

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateNEMS Residential Module Sixteen end-use services– i.e., space heating Three housing types– single family, multi-family, mobile home 34 end-use technologies– i.e., electric air-source heat pump Nine census divisions

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateSUFG Commercial Sector Model Major forecast drivers––––floor space inventoryend use intensityemployment growthenergy prices 10 building typesmodeled– offices, restaurants,retail, groceries,warehouses, schools,colleges, health care,hotel/motel,miscellaneous 14 end uses perbuilding type– space heating, airconditioning, ventilation,water heating, cooking,refrigeration, lighting,mainframe computers,mini-computers, personalcomputers, officeequipment, outdoorlighting, elevators andescalators, other

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateCommercial ModelSensitivitiesSource: SUFG 2005 Forecast

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateNEMS Commercial Module Ten end-use services– i.e., cooking Eleven building types– i.e., food service 64 end-use technologies– i.e., natural gas range Ten distributed generation technologies– i.e., photovoltaic solar systems Nine census divisions

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateSUFG Industrial Sector Model Major forecast drivers– industrial activity– energy prices 15 industries modeled– classified by Standard IndustrialClassification (SIC) system– some industries are very energy intensivewhile others are not

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateIndiana’s Industrial SectorSource: SUFG 2005 Forecast

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateIndustrial Model SensitivitiesSource: SUFG 2005 Forecast

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateNEMS Industrial Module Seven energy-intensive industries– i.e., bulk chemicals Eight non-energy-intensive industries– i.e., construction Cogeneration Four census regions, shared to ninecensus divisions

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateEnergy Peak Demand Constant load factor / load shape– Peak demand and energy grow at same rate Constant load factor / load shape for eachsector– Calculate sectoral contribution to peak demandand sum– If low load factor (residential) grows fastest, peakdemand grows faster than energy– If high load factor (industrial) grows fastest, peakdemand grows slower than energy

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateEnergy Peak Demand Day types– Break overall load shapes into typical daytypes low, medium, high weekday, weekend, peak day– Adjust day type for load management andconservation programs– Can be done on a total system level or asectoral level

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateLoad Diversity Each utility does not see its peak demand atthe same time as the others 2005 peak demands occurred at:––––––––Hoosier Energy – 7/25, 6PMIndiana Michigan - 8/3, 2PMIndiana Municipal Power Agency – 7/25, 3PMIndianapolis Power & Light - 7/25, 3PMNIPSCO – 6/24, 1PMPSI Energy – 7/25, 4PMSIGECO – 7/25, 4PMWabash Valley – 7/24, 5PM

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateLoad Diversity Thus, the statewide peak demand is less thanthe sum of the individual peaks Actual statewide peak demand can becalculated by summing up the load levels ofall utilities for each hour of the year Diversity factor is an indication of the level ofload diversity Historically, Indiana’s diversity factor hasbeen about 96 – 97 percent– that is, statewide peak demand is usually about 96percent of the sum of the individual utility peakdemands

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StatePeak Demand Capacity Needs Target reserve marginLoss of load probability (LOLP)Expected unserved energy (EUE)Assigning capacity needs to type– peaking– baseload– intermediate Optimization

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateReserve Margin vs. CapacityMargincapacity demandcapacity demandCM x100%RM x100%capacitydemand Both reserve margin (RM) and capacitymargin (CM) are the same when expressed inmegawatts– difference between available capacity anddemand Normally expressed as percentages

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateReserve Margins Reserve/capacity margins are relativelyeasy to use and understand, but thenumbers are easy to manipulate– Contractual off-system sale can be treatedas a reduction in capacity or increase indemand does not change the MW margin, but willchange the percentage– Similarly, interruptible loads and direct loadcontrol is sometimes shown as an increasein capacity

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateLOLP and EUE Probabilistic methods that account for the reliability ofthe various sources of supply Loss of load probability– given an expected demand for electricity and a given set ofsupply resources with assumed outage rates, what is thelikelihood that the supply will not be able to meet thedemand? Expected unserved energy– similar calculation to find the expected amount of energy thatwould go unmet Both are used in resource planning to ensure thatsufficient capacity is available for LOLP and/or EUEto be less than a minimum allowable level

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateCapacity Types Once the amount of capacity needed in agiven year is determined, the next step isto determine what type of capacity isneeded– peaking (high operating cost, low capital cost)– baseload (low operating cost, high capitalcost)– intermediate or cycling (operating and capitalcosts between peaking and baseload) some planners only use peaking and baseload

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateAssigning Demand to Type SUFG uses historical load shape analysis foreach of the utilities to assign a percentage oftheir peak demand to each load type Percentages vary from utility to utilityaccording to the characteristics of theircustomers– utilities with a large industrial base tend to have ahigher percentage of baseload demand– those with a large residential base tend to have ahigher percentage of peaking demand Rough breakdown:– baseload 60%, intermediate 15%, peaking 20%

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateAssigning Existing Resources SUFG then assigns existing generation tothe three types according to age, size, fueltype, and historical usage patterns Purchased power contracts are assignedto type according to time period (annual orsummer only) and capacity factor Power sales contracts are also assigned totype

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateAssigning Capacity Needs to Type Future resource needs by type aredetermined by comparing existing capacityto projected demand, while accounting forinterruptible and buy through loads, as wellas firm purchases and sales andretirement of existing units Breakdown of demand by type is notprojected to change across the forecasthorizon

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateNEMS Electricity Market Module Eleven fossil generation technologies– i.e., advanced clean coal with sequestration Two distributed generation technologies– baseload and peak Seven renewable generation technologies– i.e., geothermal Conventional and advanced nuclear Fifteen supply regions based on NERCregions and sub-regions

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateLoad Management andConservation Measures Direct load control and interruptible loadsgenerally affect peak demand but not energyforecasts– delay consumption from peak time to off-peak time– usually subtract from peak demand projections Efficiency and conservation programsgenerally affect both peak demand andenergy forecasts– consumption is reduced instead of delayed– usually subtract from energy forecast before peakdemand calculations

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateSources of Uncertainty Exogenous assumptions– forecast is driven by a number of assumptions(e.g., economic activity) about the future Stochastic model error– it is usually impossible to perfectly estimate therelationship between all possible factors and theoutput Non-stochastic model error– bad input data (measurement/estimation error)

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateAlternate Scenarios Given the uncertaintysurrounding long-termforecasts, it is notadvisable to followone single forecast SUFG developsalternative scenariosby varying the inputassumptionsSource: SUFG 2005 Forecast

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateBack to the Short Answer

ENERGY CENTERCENTERENERGYState UtilityUtility ForecastingForecasting GroupGroup (SUFG)(SUFG)StateFurther Information State Utility Forecasting Group– http://www.purdue.edu/dp/energy/SUFG/ Energy Information Administration– http://www.eia.doe.gov/index.html

State Utility Forecasting Group (SUFG) ENERGY CENTER State Utility Forecasting Group (SUFG) Time Series Forecasting Linear Trend – fit the best straight line to the historical data and assume that the future will follow that line (works perfectly in the 1st example) – Many methods exist for finding the best fitting line, the most

Related Documents:

Purdue Printing Services The School of Pharmacy and Pharmaceutical Sciences Purdue University Heine Pharmacy Building, Room 104 575 Stadium Mall Drive West Lafayette, IN 47904-2091 (765) 494-1361 (765) 494-7800 Fax www.pharmacy.purdue.edu The Purdue Pharmacist is published three times a year for alumni

Forecasting with R Nikolaos Kourentzesa,c, Fotios Petropoulosb,c aLancaster Centre for Forecasting, LUMS, Lancaster University, UK bCardi Business School, Cardi University, UK cForecasting Society, www.forsoc.net This document is supplementary material for the \Forecasting with R" workshop delivered at the International Symposium on Forecasting 2016 (ISF2016).

A Very Merry Christmas to All Boiler OUT Times, Volume 1 Issue 3, 12/12/2017 Edited by Ayo Adetunji aadetunj@purdue.edu and Yichen Fan fan151@purdue.edu Published by Boiler OUT Volunteer Program boilerout@purdue.edu Center for Intercultural Learning, Mentorship, Assessment and Research (CILMAR) Purdue University Young Hall, Room 120

92,699 Purdue alumni are members of fraternities, sororities and cooperative houses. 30% of Purdue alumni that served on either the Purdue Board of Trustees, PRF Board of Directors, Purdue Alumni Foundation Board, or boards for the colleges/schools from 2009-2019 were members of FSCL

Importance of Forecasting Make informed business decisions Develop data-driven strategies Create proactive, not reactive, decision making 5 6. 4/28/2021 4 HR & Forecasting “Putting Forecasting in Focus” –SHRM article by Carolyn Hirschman Forecasting Strategic W

Introduction to Forecasting 1.1 Introduction What would happen if we could know more about the future? Forecasting is very important for: Business. Forecasting sales, prices, inventories, new entries. Finance. Forecasting financial risk, volatility forecasts. Stock prices? Economics. Unemplo

Although forecasting is a key business function, many organizations do not have a dedicated forecasting staff, or they may only have a small team. Therefore, a large degree of automation may be required to complete the forecasting process in the time available during each forecasting and planning cycle.

Korean language learning demotivation among EFL instructors in South Korea 201 competing commitments to language learning necessitating a cost/benefit anal-ysis of the time and cost versus the perceived return on such an investment (Norton, 2013), particularly, as negative gatekeeping encounters may result in marginalization (Norton, 2000, 2001). Thus, while the notion that in a globalized .