The application of ant algorithms, as any other evolutionary optimisation method, requires a number of controlling parameters to be known a priori. These parameters are often determined by sensitivity analysis as their values dramatically affect the performance of the methods. In addition to these parameters, a penalty parameter is usually to be defined for constrained optimisation problems. An ant algorithm with a minimum number of controlling parameters is introduced in this paper for pipe network optimisation problems. This method uses the interrelation between pheromone change and initial pheromone strength to initialize the pheromone trail strength at the start of the computation. Ant algorithms with an elitist strategy of pheromone updating are known for premature convergence leading to suboptimal solutions. Such suboptimal solutions are avoided by using the concept of pheromone strength limiter introduced in the literature for TSP. The introduction of this concept, however, requires the introduction of a new parameter adding to the number of controlling parameters of ant algorithms. A sensitivity analysis was carried out to find the proper value of the newly introduced parameter. The results suggest that a value in the range of 0.15-0.3 is the best value for the examples considered. The efficiency of the proposed ant algorithm is tested against two benchmark examples in the literature and the results are presented. This method is shown to be capable of locating the best ever solutions obtained for these problems.