Energy and cost optimal scheduling of belt conveyor systems

Abstract

This work deals with the energy management of belt conveyor systems (BCS) under various demandside management (DSM) programmes. The primary objective of this work is to model the energy consumption and energy related cost of operating troughed belt conveyor systems under different electricity pricing tariffs. This research is motivated by the increasing need for energy efficiency and energy cost reduction in the operation of BCS. This is as a result of technological improvements in BCS technology leading to increasingly longer belts being commissioned and as a result of rapidly rising electricity costs. An energy model derived from established industry standards is proposed for long conveyors. The newly proposed model uses a first-order partial differential equation (PDE) in order to capture the state of material on the belt. This new model describes the conveyor's power requirement using an equation with two parameters. A system identification set-up involving a recursive parameter estimating algorithm is simulated for measurements with varying degrees of noise. The results show that the proposed model estimates conveyor power and material delivered by long conveyors more accurately than the existing steady-state models. Downhill conveyors (DHCs) are important potential energy sources that can be tapped to improve the overall energy efficiency of BCSs. A generic optimisation model that is able to optimally schedule three configurations of BCS with DHC is proposed. The economic assessment of implementing dynamic braking and regenerative drives technology on downhill conveyors is undertaken with the help of the model. The assessment shows that combining regenerative drives and optimal operation of BCS with DHC generates energy savings that give attractive payback period of less than 5 years. A chance-constrained model predictive control (cc-MPC) algorithm is proposed for scheduling belt conveyor systems with uncertain material demand on the output storage. The chance-constraints are based on the modelling of material demand by a sum of known mean demand and, zero-mean and normally distributed random component. The cc-MPC algorithm is shown to produce schedules that give a smaller number and smaller magnitude of storage limit violations compared to normal MPC and chance-constrained optimal control algorithms. An equation that gives the amount of effective storage required to meet storage constraints for a given value of standard deviation is established. The optimal scheduling of BCS under the real-time pricing (RTP) tariff is considered. This study develops a methodology for establishing the economic value of price forecasting schemes for loads capable of load-shifting. This methodology is used to show that the economic benefit obtained from a forecast is highly dependent on the volatility of the electricity prices being predicted and not their mean value. The methodology is also used to illustrate why the commonly used indices mean absolute percentage error (MAPE) and root mean square error (RMSE) are poor indicators of economic benefit. The proposed index using Kendall's rank correlation between the actual and predicted prices is shown to be a good indicator of economic benefit, performing far better than RSME and MAPE.