We present a family of simulation models of agents with energy from social interactions. We take the concept of "energy" from social network analysts Cross & Parker, from Collins's micro-sociology of interaction rituals, and from the social psychologists Ryan & Deci's studies on intrinsic motivation. We use simulation models as "tools for thinking" about what energy is, and how it relates to the take up of ideas, the formation of cultural groups and the performance of work. Our models also provide insight into phenomena from studies of "communities of practice", social capital and computer models of networks. Baker & Quinn have also developed simulations of agents with energy, and so we offer a critique of those. We develop our models as extensions of the Axelrod Cultural Model. Our family of energy models include those that ascribe "emotional energy" variously to individual agents, to agents' individual attributes, and to agents' memories of interactions rituals. Agents obtain energy payoffs from interactions based variously on their sense of autonomy, belongingness and competence. We compare the behaviour of each model and choice of payoff function through experiments to test claims derived from Cross & Parker: namely that "energisers" cause greater take up of their ideas, cause larger cultural groups to form around them, and raise the problem-solving performance of the agent population. We demonstrate this first claim for several model scenarios, but fail to find scenarios where the second two hold. We then conduct experiments to relate the capabilities of energisers to tasks of: disseminating ideas to otherwise homogeneous groups, and; spanning boundaries across cultural divides between groups. In all experiments we find two factors play critical roles in determining the diffusion and homogenisation of culture: the decay of energy charges on memories, and; the initial number of cultural traits in the population.