This chapter establishes a comprehensive methodological framework for industrial prompt engineering, examining both the evolutionary development of techniques and their strategic application across the industrial lifecycle. The intersection of Large Language Models (LLMs) with industrial applications creates substantial opportunities for accelerating innovation and problem solving in complex technical domains. However, effectively leveraging LLMs for industrial applications requires systematic approaches to prompt engineering specifically tailored to the unique characteristics of industrial contexts. The chapter presents prompt engineering techniques organized in three evolutionary generations: first-generation foundational approaches including simple prompting and few-shot learning; second-generation reasoning techniques featuring chain of thought and role prompting; and third-generation advanced problem exploration methods including tree of thoughts and self-consistency verification. An integrated strategic framework maps these techniques to industrial project phases from problem definition through implementation planning. The methodology addresses critical challenges including hallucination prevention in technical domains, safety–critical application boundaries, and organizational implementation pathways. Through systematic analysis and practical examples from petroleum refining applications, this chapter demonstrates how structured prompt engineering can transform industrial design processes while maintaining appropriate technical rigor and safety standards.