Time series lighting electricity data for rural households using Solar Nano-grids in Kenya

This data set is time series electricity use data from rural households using off-grid energy systems in Kenya. As well as indicating lighting electricity use for a real-world use case, it can give insight into active occupancy times in the mornings and evenings. This can support estimation of load profiles for higher tiers of the Multi-tier Framework for energy access by adding in load profiles for additional appliances. Two solar nano-grids (SONGs) were built in two rural communities in Kenya, as part of the Solar Nano-grids project (EPSRC ref: EP/L002612/1). One aspect of the SONGs were battery-charging systems, in which batteries could be charged at a central solar hub, and used in households to power lighting and mobile phone charging. For each battery the electricity use was recorded in real-time between July 2016 and November 2016 inclusive. The data consist of separate demand (use of battery in the home for lighting) and charging (charging at the central hub) profiles in csv files, individually for each household. The data are half-hourly measurements of average power used for the household lighting system (3 3W LED bulbs with wiring and switches). There is data for 51 households, ranging in length from 3 days to 5 months. Note that the data set is solely electricity use for the household lighting system, and does not include electricity use via the USB port that was present for charging mobile phones. The households are anonymised and are numbered in order of ascending number of days of data. The household battery packs were Li-ion with capacity 62 Wh, and the data were recorded using a FRDM K-64F mbed embedded in each. 13 post-processing steps were required to process the data gathered in raw form from the batteries into energy profiles for individual households (see reference below). These included: correcting the timestamps caused by time drift or recalibration of the RTCs, attributing batteries to the correct household, addressing logging disruptions and inconsistent logging frequencies, imposing limits on power and duration of use to remove non-representative battery use, and testing loading conditions to remove abnormal energy use. The gaps in the data and varying lengths of the data are caused by: technical challenges with the batteries, meaning that they required frequent repairing; issues with the RTC on the microcontroller being reset; difficulty in attributing data to the correct household. Between 18th July - 1st August (approx.), the charging hub was shut down and so there is a gap in all energy profiles. Graphical representations of the data for each household, and further information about the solar nano-grids project, the energy data, and the processing steps involved, can be found in Clements, A F. Data-driven approaches enabling the design of community energy systems in the Global South. DPhil Thesis. Department of Engineering Science, University of Oxford. 2019.本数据集为肯尼亚使用离网能源系统的农村家庭的时序用电数据。其不仅可反映真实场景下的照明用电情况，还能揭示家庭早间与晚间的活跃入住时段。通过补充各类家电的负荷曲线，该数据集可辅助估算能源获取多层级框架（Multi-tier Framework for energy access）的更高层级负荷分布。 作为“太阳能微电网项目（Solar Nano-grids project，EPSRC项目编号：EP/L002612/1）”的一部分，研究团队在肯尼亚的两个农村社区搭建了两套太阳能微电网（Solar Nano-Grids, SONGs）。该微电网的核心组件之一为电池充电系统：电池可在中央太阳能枢纽处完成充电，随后被带回家庭为照明设备与手机充电。2016年7月至2016年11月（含首尾月份），每套电池的用电情况均被实时记录。 本数据集以CSV格式存储，包含每个家庭独立的用电需求（家庭电池用于照明的功耗）与充电（中央枢纽充电）曲线。数据为每半小时一次的家庭照明系统平均功率测量值，该照明系统由3盏3W LED灯泡及配套布线、开关组成。数据集涵盖51户家庭的用电记录，时长跨度从3天至5个月不等。需注意，本数据集仅包含家庭照明系统的用电数据，不包含用于手机充电的USB端口的用电情况。所有家庭均已匿名处理，并按数据记录天数的升序进行编号。 家庭使用的电池组为锂离子（Li-ion）电池，容量为62 Wh，每块电池均搭载嵌入式FRDM K-64F mbed开发板完成数据记录。研究人员需通过13个后处理步骤，将从电池采集到的原始数据处理为单个家庭的用电曲线（详见下文参考文献）。这些步骤包括：校正因实时时钟（Real-Time Clock, RTC）时间漂移或重新校准导致的时间戳错误、将电池匹配至对应家庭、解决日志中断与日志频率不一致的问题、对用电功率与时长设置限制以剔除非典型的电池使用情况、通过负载条件测试移除异常用电数据。数据存在缺失且各家庭记录时长不一，其原因包括：电池存在技术故障，需频繁维修；微控制器上的实时时钟被重置；难以将数据匹配至正确家庭。大致在2016年7月18日至8月1日期间，中央充电枢纽处于停运状态，因此所有家庭的用电曲线均存在数据缺失。 各家庭用电数据的可视化图表，以及太阳能微电网项目、用电数据与相关处理步骤的更多细节，可参阅文献：Clements, A F. Data-driven approaches enabling the design of community energy systems in the Global South. 哲学博士学位论文. 牛津大学工程科学系. 2019年。