Typical traffic control and operations at work zones are not effective in mitigating the work zone bottleneck, as the common early merge behavior limits utilizing the available capacity in the closed lane. Recently, using late merge strategy is encouraged to fully utilize the closed lane, therefore increasing efficiency of the work zone. In practice, the late merge strategy does not work effectively due to the lack of compliance by drivers. This paper is aimed at exploring two work zone late merge strategies with and without enabling connected vehicles (CV) technology. This first strategy is the late merge without CV enabled. The second strategy incorporates CV-enabled cooperative merge to enhance the work zone efficiency through communications between vehicles in the open and closed lanes. The CV-enabled strategy employs a rule based decentralized control algorithm. The paper has implemented both late merge strategies in microsimulation and evaluated the operational performances. Three major performance measures, which are throughput, delay and queue length, are used in the evaluation. The results indicate that the operational performance of the CV-enabled cooperative late merge is superior to the late merge strategy without CV in most cases, especially when traffic demand is moderate, while both late merge strategies outperform the traditional early merge strategy. Moreover, effects of the control variables including the traffic demand, heavy vehicle percentage, lane distribution percentage, late merge length, reduction of gap acceptance factor, and compliance rate, have been analyzed. Ultimately, sensitivity analysis has been performed to investigate the trend of effects from the significant variables control.