The point of view taken here is that systems analysis is a kind of learning process, not data gathering, not decision making, but the production and effective communication of arguments relevant in a particular context. This idea, that the intended application of the result of an assessment has consequences for methodological choices, is beginning to spread in the LCA research community. One problem is that standard LCA methodology is developed to answer questions about environmental impacts of the current production and use of one unit of a product or minor product or process changes. When this methodology, unchanged, is used to provide answers to questions about strategic technology choice, i.e. not decisions that aim at improving a process within an existing technological environment, but with the long-term goal of changing large-scale technological systems, the result could be of little value or misleading. In many cases, LCAs produce more noise than knowledge. This observation seems to be of particular importance for LCAs of energy technologies and for how energy use is treated in all kinds of LCAs. Here, it is suggested that a better understanding of some critical methodological issues related to time, universality, cause-effect relationships, technical maturity and system innovation, could result in better studies that reveal fundamental environmental issues related to the objects of study and reduce the noise from irrelevant information. Examples are given from the technology fields of solar cells, fuel cells, batteries, renewable transport fuels and carbon nanoparticles.