Workflow technologies have become a major vehicle for the easy and efficient development of science applications. When integrating the workflow technology with the state-of-art resource provisioning technology, the challenge is to determine the amount of resources necessary for the execution of workflow. This paper introduces an algorithm named Balanced Time Scheduling (BTS), which estimates the minimum number of hosts required to execute a workflow within a user-specified finish time. The resource estimate of BTS is abstract, so it can be easily integrated with any resource description languages and resource provisioning systems. Moreover, the experimental results with a number of synthetic workflows and several real application workflows demonstrate that BTS can estimate the resource capacity close to the lower bound while the algorithm is scalable so that its turnaround time is only tens of seconds even with workflows having thousands of tasks and edges.