Sensitivity Analysis for Expansion of Transmission Systems with High Amounts of Renewables

Sensitivity Analysis for Expansion of Transmission Systems with High Amounts of Renewables

Galinas, E.

Boemer, J.C. (mentor)
Gibescu, M. (mentor)

Applied Sciences

Power Systems -EWI

SET

2013-03-28

The increasing share of Renewable Energy Sources (RES) in oday`s electricity mix has introduced new challenges for the geing European power networks. Optimal large scale deployment of wind and solar units around Europe necessitates the upgrade of the transmission system. Identifying transmission bottlenecks for fully accommodating high amounts of renewables comprises the objective of the so called Transmission Expansion Problem (TEP). However, in times of high RES in-feed and consequently high network congestion, minor RES curtailment could potentially be more economical than reinforcing another network branch. The related annualised investment costs of these transmission lines can be larger than the value of the respective RES energy integrated into the power system. Primary objective of this thesis is to perform certain sensitivity analyses in an attempt to optimise between renewable power curtailment and network reinforcement investment costs. A new heuristic approach that employs Optimal Power Flow (OPF) as its main tool, termed Least Upgrade Curtailment Sensitive Pseudo Optimisation (LUCSIPO) was developed for this purpose. The LUCSIPO approach, integrated with a statistical branch overload ranking, is a simple, straight-forward methodology that determines the least upgrades needed for obtaining a feasible solution. Its novel feature is the sensitivity on the curtailment limits applied. It is a key part that actually guides the algorithm towards a minimal cost solution, since it considers the trade-off between network upgrades and curtailment of RES power. In order to validate the LUCSIPO results, a modern Meta-Heuristic TEP approach that utilises the Ant Colony Optimisation (ACO) concept was also developed. This nature inspired algorithm, mimics the collective behaviour ants exhibit while searching for food. When an ant locates a food source, it travels back to its nest and emits a certain serum called Pheromone. Trailing this serum effectively guides the other ants towards the food source. The most visited paths will exhibit the highest pheromone concentration and thus will form the shortest ("optimum") path to the nest. When applied to the TEP, a path corresponds to a set of upgraded network branches while the path length represents the total system costs. The IEEE-39 bus New England Test System has been employed as the testing platform for the two developed approaches. Sensitivity analysis results prove the existence of an economic optimum between grid upgrades and RES curtailment. They further highlight the importance of network upgrades and high non-synchronous penetration limits for secure and cost-efficient large scale RES integration. For the selected high installed wind capacity system, the sum of curtailment and investment costs is minimised and thus regarded as the \optimum" in the at curtailment limit region between 30% and 60% of the installed wind capacity. The merit order effect, however, has a direct impact on the trade-off between RES curtailment and grid extension. The previously identified \optimum" is shifted to a curtailment limit of about 70-80% (discarding c.a. 0.75% of the available wind energy) if the operational costs are also taken into account. The LUCSIPO`s approach efficiency in providing solutions that minimise the total costs involved, has been compared against the ACO algorithm. Results show that the LUCSIPO approach introduces major calculation time speed-ups while providing solutions very close to the ACO methodology. The final goal of this thesis was to create a continental Europe High Voltage network model for potentially applying the developed approaches. Employing similar modelling techniques and assumptions as in pan-European grid integration studies, led to the development of a consistent,spatial resolution flexible DC load flow modelling platform. It integrates publicly available consumption data with externally supplied RES time series and network information for effectively modelling the European HV network.

http://resolver.tudelft.nl/uuid:31360051-943b-492d-9404-0196ea8472be

Student theses

master thesis

(c) 2013 Galinas, E.