The optimization of building operations to shift loads on the grid can be carried out using numerous flexible electrical aggregates and units assisted by different ICT solutions. Of particular importance are systems with thermal-electrical coupling such as heat pumps, chillers and co-generation power stations which use thermal inertia as virtual storage. In SGMS, two primary approaches were used: Optimized control of interruptible consumers using ripple- control units and the flexibility of buildings with the help of building automation systems. The latter are assisted by a Building Energy Agent, which acts as the communication interface to the electric power system (B2G approach). The power grid communicates using a smart grid controller.
Optimized control of interruptible consumers via ripple control
The first approach to using existing, easily accessible flexible loads was to identify electric heating and hot water systems that are outfitted with ripple controls. These units are already being switched on and off by the grid operator according to predetermined times which are strictly regulated. These flexible electric loads currently account for a total of approx. 80 MW (around 10% of peak load) in the network managed by Salzburg Netz GmbH. Using optimized triggering, a total of 10-15 MW (approx. 1.5% of peak load) can be shifted within the existing legal framework.
This kind of highly variable method for controlling the usage times of ripple-controlled loads is currently being analyzed and tested. In addition to analyzing the potential as well as the technical feasibility, the legal and regulatory framework for this kind of demand response management is being examined closely since the legally mandated control and time-of-use network tariffs  partly pose a barrier to the maximal utilization of the potential of ripple controlled load management. An area served by a primary substation with an above-average saturation of flexible loads was chosen as a test area. Here, depending on the weather, from 0.5 to 2 MW of load (up to 10% of peak load in the area) could be shifted. Customers were informed about the test in a letter, and there were no negative responses. The control actions take place according to economic criteria.
Being able to rely on existing control infrastructure significantly reduced the investment that had to be made in the test.
A disadvantage is the strong seasonally fluctuation of consumption. Furthermore, conditions within the individual households are not taken into account in the control of the units. If, for example, a central heating system had been turned off for several hours prior to a planned shut-down, comfort would dictate that the system would forgo the jettisoning of this load. The use of electric heaters and hot water systems with ripple control should therefore be viewed as a temporary solution that can be quickly implemented and less as an intelligent technology of the future.
In addition, the percentage of electric space heaters is on the decline . On the other hand, the number of heat pumps is on the rise and seems set to continue. Many factors indicate that a system of the future would be able to take the processes within the building and therefore also the needs of residents into account but would also be able to be set up in all types of buildings regardless of the type of heating or hot water system. Still, the fundamental approach using the thermal inertia of a building to flexibly manage its load will remain.
The Building to Grid approach
With the help of thermal simulation, it could be shown that even old buildings possess thermal characteristics that make it possible to shift heating loads over a period of several hours. The comfort of users is never in jeopardy since building materials react extremely slowly and most buildings also have storage units in the heating system of the building to act as a buffer.
The investment needed to configure a building for thermal modelling is normally considerable but has fallen drastically with a newly developed model that only needs the data available from an Energy Performance Certificate. The validation of these models was undertaken as part of a field trial. Adaptive systems that use sensors to conform to the conditions in a building were able to further reduce the cost and effort necessary to calibrate a model and therefore constitute a promising solution for future building systems. Such a system can be integrated in building automation. It recognizes the processes of the operator and optimizes the operation and energy of a building in light of the operator’s wishes. In order to make communication and harmonization with the electric power system possible, an additional interface is necessary.
In the Building to Grid project, building automation systems in ten existing buildings were extended to include a Building Energy Agent (BEA) which communicates with the electric power system, which is represented by a smart grid controller, fulfilling the following tasks:
- Creating energy prognoses for the components used for the thermal conditioning of the building ;
- Dispensing the load-shifting potential, having taken into account warm-up or cool-down phases and compensation for the rebound effect  within a defined period of time;
- Optimizing the utilization of on-site generated energy from e.g. photovoltaic units using electromobility and other flexible loads .
In order to supply more flexibility, thermal storage units such as hot water buffer-storage systems and the building itself are used. In total, there was a maximum flexible load potential of approx. 350 kW in the ten buildings of the field trial.
It makes sense to use a building energy agent when a building is automated or when there are several smart-grid applications. The former is however much more prevalent in commercial and industrial buildings than in residential ones at the moment. There is however predicted to be a rise in the number of in-home energy management systems (EMS) in use in households .
In order to use the potential of buildings as part of a smart infrastructure, it will be necessary in the future to include outside information in the local optimization of the building. The types of information will extend from weather data to time-sensitive information on pricing all the way to the condition of the grid and will facilitate more exact and longer-distance prognoses. The integration of different information as well as possibilities for residents to intervene in the system is shown in Figure 1.
Figure 1: Approach in the Building to Grid project (B2G)
There is a need for additional research and development in the interface between the electric power system and the building. On the one hand, the interests of grid operator, who has to guarantee security of supply at distribution level and maintain voltage at all times in addition to monitoring the utilization of the network infrastructure, conflict with those of the building operator, who wants to secure the best price on the electricity market by offering flexibility. Hence, market models or market rules at distribution level need to be created that take both sides into account and are satisfactory to both. One approach is the traffic light model in which market demands are taken into account as long as critical threshold values on the power grid haven't been reached. If these are exceeded, then the grid operator can act to stabilize the grid without taking the market into account.
On the other hand, technical details like bringing protocol standards in the interface into agreement and developing security standards for data transfer between components of the grid operator (smart grid controller) and the devices and appliances in the building (BEA).
 The tariffs are currently divided into low (from 10pm to 6am) and high (from 6am to 10pm).
 According to guideline 6 of the Austrian Institute of Construction Engineering (OIB) electric resistance heating systems are not permissible in new buildings if they are the primary system. When buildings are renovated these heaters are partly replaced by heating systems that use other sources of energy.
 In the field trials, different heating systems such as combined heat and power plants (CHP), heat pumps, electric storage heaters, and electric space heaters were utilized. These concepts are in principle also applicable to chillers and air conditioners.
 In this context the rebound effect describes the simultaneous recharging of storage units which had been switched off simultaneously during a period of high demand
 This functionality of a building energy agent is primarily used in the DG Demo Net Smart Low Volt- age Grid project
 For further information see (in German): VDI/VDE Innovation + Technik GmbH (2011): Technologische und wirtschaftliche Perspektiven Deutschlands durch die Konvergenz der elektronischen Medien. Commissioned by the Federal Ministry of Economics and Technology, Berlin