Page - 139 - in Critical Issues in Science, Technology and Society Studies - Conference Proceedings of the 17th STS Conference Graz 2018

change their routines; it is also questionable whether such a change is desirable. It would imply a decrease of knowledge about individual plants which is seen as highly relevant for good planning and operational management and might be important, for example, on the occasion of forthcoming rearrangements in the building energy system. Learning effects related to planning process Due to a low level of standardisation (no plug-in systems), absorption chiller systems require extensive individual, case-related planning. Concerning actor structures, there is a high degree of division of tasks and responsibilities. Planning must integrate contributions from several fields of competence (facility management, overall planning, electrical and hydraulic implementation planning, energy supplier, company management, building owner), and with actors with different decision-making authority and interests which are not necessarily aligned with the highest possible energy efficiency of the planned system. Additionally, data requirements differ from what practitioners perceive as “normal” (i.e., from data requirements for compression refrigeration machines). Thus, planning is associated with high communication effort and time requirements; communication problems and misunderstanding are foreseeable. Good planning results, in terms of the technological system’s optimisation towards energy efficiency, depend very much on factors effecting the quality of communication and collaboration. Good coordination, clear definition of responsibilities and sufficient time for necessary exchanges are essential but cannot be taken for granted. As involved professionals work under cost (= time) pressure and are accustomed to standard technology and procedures (which reduce time and cost), and interview partners reported an increasing practice to plan without a planner responsible for overall integration, it can be concluded that real-world conditions are opposed to optimised planning conditions in several ways. Learning effects relating to economic efficiency and perspectives Interviewees reported that pursued economic efficiency was not easy to achieve (but they – realistically – expected improvements, pointing to pending optimisations at the time of interviews). Despite recent cost reductions, investment costs are still too high (but with increasing production, further reduction is expected). Heat costs, which depend very much on local conditions (heat source, regional price structures for district heat), are crucial for economic efficiency. Additionally, in practice actors do not differentiate sharply between costs for cold systems and the costs for necessary modifications in existing technology, as budgets are limited and each expense has to be justified towards decision-makers. Not least, the estimation of cost efficiency is perceived as difficult, as there are different models and actors were not secure about “best” or appropriate models. In order to reduce costs, actors demand more standardised products. Following the researchers’ statements, this is hardly possible. Conclusions Interviews with practice actors show that and how field tests for innovative absorption chiller systems result in comprehensive learning processes, comprising different aspects from user context to legal framework conditions. These results confirm and refine outcomes of the constellation analysis (see Keppler 2018) by more detailed insights on actor constellations, 139