Wearable devices are having a transformative impact on personalised monitoring and care. However, they frequently have limited battery life, requiring charging every few days; a major source of user frustration. Kinetic energy harvesting may help overcome this, collecting energy from the user's motion to allow the device to self-charge. While there are many works which have investigated wearable energy harvesting potential, none have incorporated socio-economic factors which affect activity, such as occupation type, on energy harvesting potential. We use the UK Biobank free-living accelerometer dataset to investigate the impact of occupational patterns on energy harvesting potential for the first time. We identify that those following shift patterns have a different distribution of when power is available, with those who work shifts having the most power intense period spread over a longer period of the day compared to controls. When stratifying into day or night shift work, we identify that those who work night shifts have a large variation between participants, as their most energy dense period is spread over the entire 24-hour period. This is compared to day shift workers who have the most power concentrated within a substantially smaller window, typically in the morning. Considering these socio-economic factors may affect system design of wearable energy harvesters.