The feasibility of district cooling systems is linked to their efficiency, which is associated with their capital and operational costs. For reasons of ownership and maintenance, the piping network commonly follows the city's street layout. This paper examines the interdependencies between the street grids and the efficiency of district cooling systems, measured by five cost indicators, using Singapore as an example. We create a parametric model of street grids in Grasshopper featuring three input variables: block area, block elongation, and site area. The boundaries of these variables and other urban design parameters are based on a survey of the blocks in five high-density areas in Singapore. We generate hundreds of street grids using a quasi-Monte Carlo Saltelli sampler. For each sample, we use the City Energy Analyst, an urban energy modeling platform, to assess the costs of a hypothetical district cooling system. Using Sobol' sensitivity analysis, we determine the contribution of each variable to the five cost indicators of the district cooling system. Our results show that the block area has the highest impact on the cost indicators, followed by block elongation. We then discuss the implications of these findings for the design of street grids in high-density urban areas. Urban designers may apply these findings when designing street grids for efficient district cooling systems in high-density cities.