Dependence of distributed generation (DG) outputs and load plays an essential role in renewable energy accommodation. This paper presents a novel DG hosting capacity (DGHC) evaluation method for distribution networks considering high-dimensional dependence relations among solar radiation

wind speed

and various load types (i.e.

commercial

residential

and industrial). First

an advanced dependence modeling method called regular vine (R-vine) is applied to capture the complex dependence structure of solar radiation

wind speed

commercial loads

industrial loads

and residential loads. Then

a chance-constrained DGHC evaluation model is employed to figure out maximum hosting capacity of each DG and its optimal allocation plan with different operational risks. Finally

a Benders decomposition algorithm is also employed to reduce computational burden. The proposed approaches are validated using a set of historical data from China. Results show dependence among different DGs and loads has significant impact on hosting capacity. Results also suggest using the R-vine model to capture dependence among distributed energy resources (DERs) and load. This finding provides useful advice for distribution networks in installing renewable energy generations.