The construction and application of a multiple nitrogen liquefier unit (NLU) optimal scheduling tool is discussed. Constrained by customer demands and subject to electricity spot market prices over a week-ahead horizon, a retrospective optimiser (RO) determines the minimum scheduling costs. Plant start-up penalties and inter-site optimisation capabilities are incorporated into the optimisation model to emulate realistic operational flexibilities and costs. Using operational data, actual process schedules are compared to the RO results leading to improved process scheduling insights; such as increasing afternoon NLU operation during the spring to utilise lower power pricing caused by high solar generation. The RO is used to output a trackable load management key performance indicator to quantify potential and achieved scheduling improvements. Subsequently, correlations between renewable energy generation and spot market power prices are developed. Forecast pricing is used within a predictive optimiser (PO) to automatically generate an optimal schedule for the week ahead to meet projected customer demands. The RO provides potential hindsight savings of around 11%, and the PO up to 8% (representing significant cost savings for such energy intensive processes).

讨论了多氮液化器优化调度工具的构建和应用。受客户需求的约束，并且受一周前的现货市场价格的制约，回顾性优化器（RO）确定了最低调度成本。工厂启动处罚和站点间优化功能被合并到优化模型中，以模拟现实的操作灵活性和成本。使用操作数据，将实际流程计划与RO结果进行比较，从而改善流程计划的洞察力；例如在春季增加下午的NLU操作，以利用由高太阳能发电量产生的较低电价。RO用于输出可跟踪的负载管理关键绩效指标，以量化潜力并实现调度方面的改进。随后，开发了可再生能源发电与现货市场电价之间的相关性。在预测性优化器（PO）中使用预测价格来自动生成未来一周的最佳计划，以满足预期的客户需求。RO可以节省大约11％的事后观察力，PO最多可以节省8％（这意味着此类能源密集型过程可节省大量成本）。