ebook img

Naresh Acharya, GE Global Research PDF

38 Pages·2014·1.92 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Naresh Acharya, GE Global Research

Fast Dynamic Simulation-Based Small Signal Stability Assessment & Control Dr. Naresh Acharya GE Global Research [Collaborating Organizations: GE Energy Consulting, PNNL, SCE] June 18 2014 1 Context/Vision Today’s Practice Proposed Paradigm Presentation Outline  Project Objectives  Technical Approach  Parallelization of PSLF  Alternative DAE Modeling  Fast Contingency Screening  Oscillation Damping Control  Technical Accomplishments  Project Team 3 Project Objectives  Apply advances in high performance computing techniques to develop fast Contingency Screening and Control Action Engine (FSCAE) for proactive small signal stability assessment, prediction and control. Develop mathematical and high-performance computing (HPC) techniques applicable to power system fast dynamic simulation. Implement HPC techniques in power system dynamic simulation software Develop fast contingency screening method Develop oscillation damping control method Verify and validate speed enhancement of dynamic simulation and decision making methods. 4 Technical Approach Today’s practice Proposed approach Contingencies C1 C2 C3 C4 Contingencies C1 C2 C3 C4 n u A A m s c A e s c d se le Processors P1 P2 P3 P4 Processors P1 P2 P3 P4 cir av s e a n m ar s l c e t o e n de Time (mins.) 10 9 8 9 Time (mins.) 3 2 2 2 vl d t e r Minimum time to make decision -> 10 minutes Minimum time to make decision -> 3 minutes s Contingency Screening Proactive Decision Reduce contingency space Proactive guidance to operators C P to ensure system stability a r p e d a Estimation in frequency domain b ci tili vit & validation in time domain Control based on operating y e using HPC point adjustment 5 Parallelization Approaches - Pros & Cons Approach 1: Use Approach 2: Alternative parallelization under DAE modeling (implicit current PSLF architecture integration) • Pros • Pros • Faster to implement • Expect greater speed gains than approach 1 • Less code changes • Can be used as the basis for • Cons the development of other • Speed gains are limited by tools (small signal analysis) the speed of slowest loop on • Cons current program architecture • Slower to implement • Larger changes can be • More code changes and more challenging and require code development significant modifications • Need to reformulate solution approach in PSLF 6 Parallelization of PSLF Challenges faced  Program is already well written and optimized  Complex program structure and legacy code (program has been written over 30 years)  Overhead costs of parallelization methods will adversely impact performance on small cases  Replacement of linear solver involves significant changes in the core 8 Selecting an appropriate solver for the problem at hand  Literature review (and preliminary results) have indicated that for current power system power system matrix sizes  Direct methods have superior performance over iterative methods  Serial solvers are faster than parallel ones  As problems grow larger, iterative methods are expected to outperform direct methods 9 PSLF core architecture improvements  The most effective way to reduce the solution speed of PSLF dynamics is a combination of two techniques  Parallelization of ODEs  Fast linear solver (Network)  Solver speed will be directly dependent on matrix sparsity structure and problem size 10

Description:
Serial solvers are faster than parallel ones. ➢ As problems grow Preliminary results. ➢ Execution time of a 1s simulation on an real size system . using full admittance matrix with implicit integration . Dr. Santosh Veda. Dr. Juan
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.