What i am basically trying to do is to acquire optimum load flow solution of power grid of a certain area. representing every Generating stations and load centers makes the system bulky and the jocobian matrix would be massive to solve. I am seeking for standard procedure in which i might be able to develop simplified equivalent power system network to work with.
We have published a few papers on similar analyses recently (see my publications with Seyed Mostafa Safdarnejad as primary author). First, most of the important grid issues are inherently transient, so only a dynamic equivalent model would be appropriate. Power grid systems have many characteristics that are not common in equivalent grid systems, including random intermittent sources (solar and wind), dynamic load changes, time-constrained responses (NG CS, coal and nuclear respond slowly, NG simple cycle and hydro respond rapidly), scheduled outages, significantly varying energy retail prices and generation costs, etc. There are also forcasting issues (primarily for wind and for demand) and unscheduled shutdown issues. Finally, there are significant load limits and turndown ratio limits. Many of these things could not be very accurately represented by traditional equivalent circuit analyses.
Bishal,
I hope that I understood your question properly. So, you want to develop an equivalent power system model, which represents a much larger power system, correct?
If the answer is yes, of course, you can do that. You can cut the larger power system into any area you like. Then, you will have internal area and external area. Internal area is something you have interest and external area is something you have no interest. Then, try to model power injection or equivalent load at the boundary buses. Those injection and withdrawal will represent power flow between internal and external areas. Doing this way will give you equivalent power system model. I am not sure exactly what you are trying to achieve as your end-goal.
Or do you want your new power system a little bit differently in the sense that say your original power system has 10,000 buses and messy network, but you want to represent it as 50 buses and less-messy network like 9 bus system or 14 bus system. If that is your goal, I would say, this is nearly impossible. You can NEVER make a very large system of 10,000 bus as equivalent to 14 bus system. Some researchers did that but it's just terrible!!
Let me know if I clear your question enough or if you have additional questions. I'll be more than happy to answer them.
Regards,
Jeremy
I agree with the other two replies. The extent of reduction of a power system is study dependent. I am not aware of a method which will reduce any power system to a certain number of buses.
Dear sir Larry, thanks for help. Will definitely look into your paper in detail.
Dear sir Jeremy and Abu Hamed,
you are spot on. To be specific, i am considering to simplify the national grid of my country which comprises of 12-13 major generation stations and 10-12 major load centers. that makes around 50-55 buses considering every major load dispatch centers and substations as well.
I love the idea of developing the internal area and external area but i am not sure if i could make the buses in the boundary to represent the dynamics of multiple external buses all together.
Also, sir Abu Hamed,
i am not myself sure of any specific methodology here. your expertise in this matter would be of great help.
Note: My end goal would be to extract the optimum load flow solution of the power network and implement it using FACT devices.
Theoretically, splitting the larger network into zones and finding the Thevenin equivalent for each zone is an ideal approach and representing that by a source and equivalent impedance. That can be aggregated into your designated number of buses and loads. But that is being simplistic. If you have complex power sources and intermittent loads, I presume the resulting network may not be very accurate and representative but worth the analysis. There are other methods for reducing a large power network however. It may be useful to identify the EHV or HV backbone of the network with the respective major transmission substations; and lump MV and LV lines and loads together to have a reduced model. We have used this method when analyzing a large power pool and results are reasonable good. Unfortunately, that work has just been completed successfully and results have not yet been published for me to point that out to you for consideration. Best wishes.
Bishal,
First I thought you are referring to a system with 10,000 buses. Now, you said, it would have about 50-55 buses. That is not large at all and you should be able to model them accurately the way you want (including FACTs devices). Are you aware that some IEEE test system has 118 bus, 300 bus, which are much larger than what you described? There is one IEEE model with over 1000 bus. So, your system is NOT large at all and you should model them accurately.
If you have an OPF code, you just have to expand some dimensions and that will do the job!!
Let me know if you need help. Send email to me: [email protected]
Regards,
Jeremy
Thanks Jeremy sir,
Would definitely seek for your help via mail now onward if that is all right.
Regards,
Bishal
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India