Climate change is causing profound changes in energy and power systems around the world. Although many papers have confirmed the impact of environmental weather on electricity demand, the complex nexus of regional environmental temperature and electricity demand instability remains incomplete, especially neglecting the understanding from the perspectives of multi-quantiles. This paper uses a unique dataset of 15-min the high-frequency grid load situation of Datong, a northern city in China, to construct a comprehensive indicator to measure the regional electricity demand instability. Then, employing the Quantile-on-Quantile approach, this paper evaluates the association between intraday temperature difference and both the composite and multidimensional electric instability indicators. According to the empirical analysis, a larger environmental temperature difference always rises the instability of the electricity demand, which is stronger when both two variables are at higher quantile states. There are heterogeneities among the effects on multi-dimensional indicators of electricity demand instability. Environmental temperature difference provides a negative influence on the load rate, while it will magnify other instability indicators. Furthermore, we also assess the impact of daily extreme temperatures and divide the realized volatility of load into two directions for discussion. Our research presents important knowledge about responding the climate change-induced electric system risk.