I have been told that on a previous lab project my client saw a 20% reduction in peak cooling using recovered condensate for pre-cooling. The 20% is based on collected data, so it was not modeled. This was on a large DOAS. I need to quantify the reduction in peak cooling and annual energy consumption.