Click to edit Master title style
Edit Master text styles
Click to edit Master title style
Edit Master text styles
Click to edit Master title style Click to edit Master title style - - PowerPoint PPT Presentation
Click to edit Master title style Click to edit Master title style Edit Master text styles Edit Master text styles Introduction to TESLA Forecasting Solutions Who are TESLA Forecasting Solutions? Experts in electricity and gas Not to be confused
Edit Master text styles
Edit Master text styles
Edit Master text styles
Edit Master text styles
Experts in electricity and gas demand forecasting since 1992 – this is our niche Helping clients across the globe with offices in the US, UK and New Zealand Not to be confused with Tesla Motors
Edit Master text styles
Edit Master text styles
Edit Master text styles
Edit Master text styles
TESLA
Products
Short Term Long Term Weather Risk Weather Decomp
Weather Data
Load & Customer Counts
Edit Master text styles
Edit Master text styles
MeteoGroup provide weather forecasts, updated every hour, for 14 day horizon Typical model contains several hundred hourly weather variables, not just temperature TESLA can utilise comparable weather forecasts as required
Powered by
Edit Master text styles
Edit Master text styles
MODEL
Winter and summer model Time varying coefficients School holiday and vacation patterns Interaction
variables Lag structure for delay effects Latent variable to proxy structural change
Edit Master text styles
Edit Master text styles
Up to 14 day ahead forecast Can be updated every hour Simple risk analysis using the TESLA “Temperature Sensitivity” tool Short Term Forecast
Edit Master text styles
Edit Master text styles
Same model - longer term scenarios consistent with short term Weather Risk solutions used after 14 day horizon Evidence based trends from demand data incorporated into the future Long Term Forecast
Edit Master text styles
Edit Master text styles
Weather risk utilises historic observed weather Many scenarios computed and ordered into percentiles Percentiles are customisable dependant
Useful for risk and scenario analysis Weather Risk Scenarios
Edit Master text styles
Edit Master text styles
Adjusts historic demand series to climatic average “normal” weather conditions Hot summers and cold winters distort the underlying demand trends Year-on-year change can indicate network growth or shrinkage Useful for network and financial planning Weather Decomposition
Edit Master text styles
Edit Master text styles
Awarded 2nd place out of 100+ participants in TEPCO forecasting challenge First time forecasting load in Japan, with minimal weather data The only overseas firm to place TESLA and TEPCO began collaborating after the contest
Edit Master text styles
Edit Master text styles
Golden Week Forecast
1.06% 1.11% 1.56%
0% 1% 2% Children's Day Observed Mon 6th May 2019 Children's Day Sun 5th May 2019 Greenery Day Sat 4th May 2019
8am Day Ahead MAPE
Edit Master text styles
Edit Master text styles
Tony Baker
Asia Pacific Regional Director
3/34 Te Pai Place Henderson Auckland, 0610 New Zealand +64 (0) 9 551 5039
tbaker@teslaforecast.com
www.teslaforecast.com
Mark Todoroff
Business Development Director
3/34 Te Pai Place Henderson Auckland, 0610 New Zealand +64 (0) 9 551 5039
mtodoroff@teslaforecast.com
www.teslaforecast.com