NEW TOOLS TO ENCOURAGE ACTIVE DEMAND RESPONSE IN COMPETITIVE 
ELECTRICITY MARKETS: IMPLEMENTATION AND SETTLEMENT


     In the new situation where electricity markets are involved nowadays, competition is becoming more 
and more important for the power sector. In this competitive framework, the different technical and 
economical mechanisms to achieve an optimal management of electricity systems must be taken into 
account. For that reason, the active participation of consumers in electricity markets-defined as the ability 
of a customer to react to energy prices has to be considered as a fundamental objective to make 
deregulated markets work properly. 
     Some years ago (in the mid 70s), different programs began to be developed in order to increase active 
participation of demand; the expected results were not achieved though. Some of the explanations behind 
this fiasco are: from a technical point of view, more accurate systems that permit the acquisition and 
processing of information were vital; logistically, it is important to show these programs to the customer in 
a clear detailed and attractive way, as well as the definition of a clear mechanism for an objective payment 
the customer who provides with this service, based on price signals.
     This dissertation is aimed at designing and implementation of tools to promote the use of active 
demand response for the management of electricity systems. In order to achieve this objective, different 
actions have been performed, including the design and implementation of a system for the control and 
management of facilities which makes it easier for the active demand response use. Thus, after analyzing 
the limitations of existing control systems for this purpose, a new system is proposed. It permits one to get 
in touch with different agents who may be interested in the use of available distributed energy resources.
     Moreover, this system represents an essential tool to establish a mechanism to measure the actual 
response of demand in order to facilitate the process of payment to consumers for their participation in 
operation markets. It is essential to establish an accurate methodology to compare the real consumption 
during the control action with the energy that the customer would have consumed in case any control 
action had not been carried out. This means that, on one hand, an accurate measurement of the 
consumption must be provided by the proposed system and, on the other hand, a reliable and accurate 
forecast of what the customer would have consumed if no actions had been performed have to be 
obtained. Due to the low aggregation level in the considered loads, a new prediction method based on 
artificial neural techniques has been developed. This method is proposed for the prediction of 
consumptions for each process or end-use of energy. Here, the training of the neural networks is done by 
using real measurements or simulations obtained by means of physical models for the different processes.