The loss of sorbent CO2 capture capacity and the extent of particle attrition over iterated cycles are relevant to the design of calcium looping processes. Thermogravimetric analyzers or single batch fluidized bed reactors are typically used at the lab-scale to evaluate the sorbent performance. One drawback of these devices is that they do not reproduce the thermal history that is actually experienced by sorbent particles in real looping cycles. In this study, a novel experimental device is proposed to overcome this limitation. The apparatus consists of two interconnected batch fluidized bed reactors operating as calciner and carbonator, respectively. The two reactors are connected to each other by a duct (whose openings can be located at an adjustable level above the gas distributor) which permits rapid and complete pneumatic transport of the sorbent (limestone) between the reactors. Preliminary tests at ambient temperature were first performed to find the best operating conditions in terms of limestone conveying efficiency. Under the optimal experimental conditions > 95% collection efficiency of the limestone was obtained. High-temperature tests were carried out to simulate multiple calcination/carbonation cycles. The results indicate that the device is able to effectively transfer the sorbent between the two reactors while keeping the temperature variations limited. The test reactor has been used to assess the effect of the thermal history of the limestone on its CO2 capture capacity and attrition tendency. The results were compared to those previously obtained with the same limestone under comparable operating conditions in a single bed apparatus. The comparison reveals that the thermal history experienced by the limestone has a non-negligible effect on its performance. (C) 2016 Elsevier B.V. All rights reserved.