Hi all, I have monthly load how can I extract daily or hourly load from it?
Monthly load profile is an aggregate of all the days in that month. You cannot deduce daily load profile from it.
You may get average value. If monthly load is 300 units (1 unit = 1kWh), daily average load is 300/30 = 10 units. Similarly hourly consumption.
As everyone is saying, there is no definite answer of your question. It is difficult to deduce realistic daily profile from monthly profile. However, you can follow the way Shady said.
you could use for example standardized load profiles (SLP) which are avaible on the websites of most transmission system operators (in hourly or 15 minute values). There are different types available, e.g. private homes, industrie etc. (I attached a SLP for private homes from the German TSO 50Herz in 15 minute values). The SLPs are in gerneral normalized to 1000 kWh/a. Then you could adapt theses SLPs to reach your monthly consumption. This will not be the exact load profile but at least you will have the daily and weakly characteristics.
Dear Abdullah,
I entirely agree with Alireza. You can only determine this load profile through a bottom-up approach and not top-down one.
See the following manuscript.
Regards,
Morteza Shabanzadeh
http://mortezash.blogfa.com/
Data Seasonal, Weekly, and Daily Components Extraction of Hourly ...
Monthly load (average or peak) you cant get the corresponding hourly or daily peak load and you can get the average value for the hourly or daily load.
Abdulla sent me a graph with montly aggregated min/max values. I think that Abdulla would like to generate variables with underlying values for each period's min/max values? If so, are you assuming that any given value represents the mean of that distribution? So for example, 5560 was the first max value (month 1). Do you want to generate X number of values with a mean of 5560? If so, what number of values do you want to generate, and what is the distribution you expect (ie., mean of 5560 and a variance of what?). This can look differently depending on distribution (normal, Poisson, etc.). Also, you may want to consider the covariance between successive periods if there is reason to believe that there is autocorrelation.
It is almost a very Difficult task.
But you may think of interpolating,using curve fitting Technique, first for monthly to daily,then from daily to hourly in different stages in terms of 8 hrs. slots and so on. Best will be a Cubic curve Fitting, The Data may be only approximate and may not reflect sudden changes between two consecutive days due to weather Change. There are Load forecasting Methods considering weather,
Refer: Chapter 4
Curve Fitting
Link: http://www.mathcs.emory.edu/~haber/math315/chap4.pdf
P.S.
You may also take a set of daily or hourly data and see whether this curve you have chosen fits fairly well with this data. If so you may use that curve for interpolation from your monthly data.
you can try curves with polynomials of a different degree.
This is a cumbersome Procedure and still we are not 100% confident of the Results we gt. It is a Rough Approximation.
P.S.
See
http://cired.net/publications/cired2015/papers/CIRED2015_0107_final.pdf
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts