Carpool lanes that are not filled to capacity can spell trouble, particularly when deployed on narrow freeways with only 3 lanes in each travel direction. Two alternative treatments for underused carpool lanes were evaluated with this in mind. Both can boost discharge flows by opening a portion of the carpool lane to all freeway traffic in the vicinity of a bottleneck. The simpler of the two alternatives opens the lane segment on a permanent basis. The other does so intermittently. Both alternatives leave the carpool lane intact on segments well upstream of an active bottleneck. Carpools are thus still enabled to bypass slower-moving queues in a freeway’s regular-use lanes. Both alternatives have previously been proposed, but were yet to be tested.

The present tests entail simulations of a real 3-lane freeway. The simulation model was calibrated to roughly match unfavorable conditions presently triggered by the site’s underused carpool lane. Experiments were thereafter performed by varying travel demands parametrically. These tests unveiled why the alternatives are likely to be better options than rescinding the site’s carpool restriction and opening the full length of the lane to all traffic. Simulations also show that, unless the carpool lane is nearly filled to capacity, the alternatives perform better than does the traditional practice of running the lane through a bottleneck. The simpler of the two alternatives performed well when the carpool lane operated at less than two-thirds its capacity. The intermittent treatment was found superior when carpool demands were higher. The savings in Vehicle Hours Traveled (VHT) and People Hours Traveled (PHT) generated by the alternatives were often large. Deploying the simpler alternative under the site’s present-day conditions, for example, was found to reduce PHT and VHT by more than 1,000 (hours) each, over a 5-h period spanning the afternoon rush. These improvements did not come at great cost to carpools. To the contrary, carpools often benefited from the alternative treatments, despite having at times to traverse the site’s bottleneck sans priority. This is because by diminishing regular-lane queues, the alternatives also diminished the deleterious friction effect on carpool-lane speeds. Policy implications for carpool- and other kinds of special-use lanes are discussed.

未满载的拼车车道会带来麻烦，尤其是在狭窄的高速公路上部署时，每个行进方向只有3条车道。考虑到这一点，对未充分利用的拼车车道的两种替代处理方法进行了评估。两者都可以通过在瓶颈附近为所有高速公路通行打开一部分拼车车道来提高排放流量。两种选择中最简单的一种将永久性地打开车道线段。另一方则间歇性地这样做。两种选择都使拼车车道在活动瓶颈上游的路段完好无损。因此，拼车仍然能够绕过高速公路常规使用车道上行驶缓慢的队列。两种替代方案先前都已提出，但尚未进行测试。

目前的测试需要对真实的3车道高速公路进行仿真。对该模拟模型进行了校准，以大致匹配该站点当前未充分利用的拼车车道引发的不利条件。此后通过参数化旅行需求进行实验。这些测试揭示了为什么替代方案可能比取消站点的拼车限制并向所有车辆开放全长的车道更好。仿真还表明，除非拼车车道几乎满载，否则替代方案的性能要比传统的通过瓶颈运行车道更好。当拼车车道的行驶能力不足其通行能力的三分之二时，这两种选择中的一种较为简单。当拼车需求更高时，发现间歇治疗效果更好。替代方案所节省的车辆行驶小时数（VHT）和人员行驶小时数（PHT）通常可以节省很多。例如，发现在站点当前条件下部署更简单的替代方法可以在整个下午高峰期的5小时内将PHT和VHT分别减少1000多个小时（小时）。这些改进并没有给拼车带来很大的代价。相反，拼车通常可从替代处理中受益，尽管有时不得不绕过该站点的瓶颈而没有优先考虑。这是因为通过减少常规车道队列，替代方案还减少了对拼车车道速度的有害摩擦影响。讨论了拼车和其他类型的特殊用途车道的政策含义。