ECE 562 - Spring 2020

Project 1 - Due Feb 4

You may use any power flow software you wish. However, if you don't use MATPOWER, then make a note on your project report describing the software that you used (e.g., PTI's PSS/E version xx, attached C code, attached Excel spreadsheet).

Generator Unit Outage Effects on Branch Loading

1. Choose any generator unit except the swing. Assume that the swing is the only pickup generator. I recommend a generator that is relatively far away from the swing, otherwise the questions at the end may be hard to answer.
2. Estimate the post-contingency branch MW loadings via the Power Transfer Distribution Factors (PTDFs). PTDFs are defined for all buses. Generator Shift Factors (GSFs) are the same as PTDFs at the generator buses. Note: the zip archive contains a MAT file for MATLAB, and an ASCII file. Both files have a 78x64 table of PTDFs. After unzipping the archive, you can use "load ptdf_esca64" to import the "a" matrix containing the PTDFs as formulated in lecture. (Actually, GSFs were formulated in lecture, but PTDFs use the same process.) I decided to include the swing (bus 59) column, even though it's all zero.
3. Calculate the actual post-contingency branch MVA loadings via Newton-Raphson power flow "runpf('esca64_n')" with the generator out of service. Note: use the "from-end" of the branch to measure the MW and MVAR flow, even though the "to-end" might have higher flow in some cases. In general, this is not appropriate, but it will make the project simpler and allow you to focus on the most important aspects of GSFs.
4. Compare your GSF based branch loadings to your actual NR power flow based branch loadings and submit the deliverables below.

Deliverables

1. Brief description of project.
2. Short description of your simulation, including solution summary (e.g., Newton-Raphson power flow converged in X iterations), list of active control devices (e.g., Qgen limits enforced, etc.), generator unit outaged (bus number, Pgen, Qgen).
3. Table with the post-contingency top 10 most heavily-loaded branches (ranked by loading percentage -- ignore the direction) based on the actual loading determined by power flow. Your table should include the following columns:
   1. Branch data (From, To, MVA rating, pre-contingency MVA loading)
   2. PTDF
   3. Estimated post-contingency MW loading from PTDF calculation
   4. Actual post-contingency MVA loading from power flow
   5. Estimated post-contingency percentage loading from PTDF calculation
   6. Actual post-contingency percentage loading from power flow
   7. Post-contingency percentage loading error: estimate - actual (e.g., 84%est - 87%act = -3% error)
4. Table with the post-contingency top 10 largest magnitude estimation errors (ranked by percentage loading error). Rank them based on the magnitude of the error, but include the sign so that it's clear which percentage loadings were underestimated and which were overestimated. Your table should include the same ten columns as the previous table.
5. Brief discussion of your results, including answers to the following questions:
   1. Do GSF-based calculations accurately predict the actual branch loadings? Why? Why not?
   2. Are there any significant errors in the GSF-based estimates of branch loading levels? Where?
   3. What do you think causes the errors, if any?
   4. Does the location of a branch relative to the generator unit outage affect the errors? How?
   5. Did you notice anything else?
6. All students must submit the code (MATLAB, C, Excel spreadsheet, etc.) they used to calculate the post-contingency branch flows as a section within the report. Then, the report file must be submitted via SafeAssignment on Blackboard. Main Campus students also must submit a hardcopy report at the beginning of class.