This paper presents two new mixed-integer linear programming (MILP) models for more flexible operation of a combined heat and power (CHP) plant connected to a district heating load by a pipeline. The optimization models consider the heat-storage capabilities of the pipeline to better react to volatile energy prices and heat demands, varying both the flowrate and temperature of the hot water coming out of the CHP plant. They rigorously account for varying time delays, in a simpler (linear) way than previous works. The new scheduling formulations achieve a very good computational performance by relying on either a discrete-time or hybrid discrete/continuous-time representation, where all time slots of the grid are preassigned to a given period with constant price and demand (e.g., an hour of the day). Results for a motivating example inspired by a German system show additional revenues of 345 k(sic)/year for a 10 km pipeline with a diameter of 0.7 m (3850 m(3)), compared to operating the pipeline at a constant temperature. This value is strongly dependent on the storage capacity, increasing to 141S k(sic)/year for a volume of 10,000 m(3).