Smart Transformation of Ningbo’s Petrochemical Sector: Beyond Oil Prices as the Sole Variable

Oil prices are no longer the only variable: The smart transformation of a petrochemical city

On April 10, 2026, at the Zhenhai Refining and Chemical Company in Ningbo, technicians were overseeing operations at a 3,000 cubic meter spherical tank while using control panels. Inside, wall-climbing polishing and magnetic powder detection robots were working in coordination. According to Tang Quming, the Mechanical Power Department Manager at Sinopec Ningbo Zhenhai Refining Company, the use of robots has significantly reduced inspection times. Previously, scaffolding needed to be erected for manual inspections, which could take 5 to 6 days for a single tank. With robotic operations, this step is eliminated. Given that there are 112 spherical tanks at the Zhenhai Refinery, this translates to a savings of over 500 days, minimizing downtime for inspections and consequently reducing costs.

Guo Weican, Chief Engineer of the Zhejiang Special Equipment Research Institute, noted that with the introduction of polishing robots, operators can now work from outside the tanks, enhancing efficiency and safety. Inspections that once took 5 to 6 days can now be completed in just a day and a half, while significantly lowering labor intensity. This is a snapshot of the intelligent transformation occurring within the Ningbo petrochemical industry, which is not only a local technological upgrade but also a critical case study in observing economic trends in coastal industrial cities amidst global energy fluctuations.

The volatility in oil prices due to events in the Strait of Hormuz first impacts the global commodities market and is quickly reflected in the economic data of China’s coastal industrial cities. As a manufacturing city, Ningbo’s petrochemical sector is a cornerstone of its industrial base. Therefore, accelerating digital transformation is essential to promote intelligent, green, and integrated upgrades within the petrochemical industry. By adopting artificial intelligence (AI) technologies and digital methods, Ningbo seeks to explore new pathways for process-oriented industrial development in the smart economic era.

On March 30, a digital transformation supply-demand matching event for the petrochemical industry took place in Ningbo, marking a rapid acceleration in the transformation progress of the petrochemical sector in both Ningbo and Zhejiang Province. Qian Feng, an academician at the Chinese Academy of Engineering, emphasized that China has moved beyond the information phase and is now in the era of digitalization and intelligence. He stated that it is important to consider how AI can enhance manufacturing capabilities. “How can AI help us build an industrial brain that can command the most precise operations in process-oriented industries?” he asked, stressing the need for AI to better understand business logic and behave intelligently.

Zhao Lujun, Vice President of the Marketing Center at Control Technology Co., Ltd., explained that while AI has immense potential in the industrial sector, the challenges of implementation are significant due to the high barriers of entry and the complexity of data. The constraints of the “three transmissions and one reaction” (referring to momentum transmission, heat transmission, mass transmission, and chemical reactions) complicate the integration of AI into industrial processes. However, successful implementation could lead to precise predictions, optimized production, and reduced carbon emissions, addressing challenges that traditional digital tools struggle with.

In Ningbo, Control Technology has initiated multiple smart application collaborations with Zhenhai Refining and other leading petrochemical enterprises, allowing many innovative industrial AI scenarios to materialize. Nevertheless, Zhao cautioned that AI is not a panacea. Companies need to clarify their problems and target models while considering their human resources, technology, and operational modes to develop a systematic plan, rather than following trends blindly.

In 2025, Ningbo’s economic growth was affected by falling prices of chemical products. That year, the city’s industrial output increased by 5.3% year-on-year, but the producer price index for industrial products dropped by 3.3%, indicating that while Ningbo maintained industrial stability, the impact of price cycles weakened profitability. In response, Ningbo accelerated its digital transformation in the petrochemical sector. In September, a digital transformation scenario map for the petrochemical industry was launched at the Ningbo Intelligence Expo.

By 2026, as new changes emerged in the international energy market and upstream oil prices stabilized, Ningbo’s petrochemical enterprises found themselves in a relatively advantageous position for industrial upgrades. The city’s economy is particularly sensitive to fluctuations in petrochemical prices due to the significant role of the petrochemical industry within its industrial framework.

Tu Jiong, Deputy Director of the Ningbo Economic and Information Technology Bureau, indicated that as one of China’s seven major petrochemical bases, the green petrochemical cluster in Ningbo is a key focus for creating a trillion-yuan industrial cluster. It is also one of the first recognized national advanced manufacturing clusters in the field of green petrochemicals, characterized by a complete industrial chain led by refining and ethylene, as well as being the world’s largest polyurethane production base, with an industrial output value nearing 550 billion yuan by 2025.

Ningbo hosts large-scale leading enterprises in process industries like Sinopec Zhenhai Refining, while numerous small and medium-sized manufacturers are spread across sectors such as automotive parts and equipment manufacturing. This dual structure means that Ningbo’s economy is particularly responsive to fluctuations in upstream commodity prices. When industries such as petrochemicals experience cyclical changes, the industrial climate of the city often responds rapidly. This is not only a challenge faced by Ningbo but is also a common issue for many industrial cities dominated by the petrochemical sector.

Qian summarized this issue by stating that while leading companies in China’s petrochemical industry possess scale advantages, they have yet to establish core competitive strengths that provide significant bargaining power. The goal of creating a world-class green petrochemical sector is continuously pursued in Ningbo.

The Ningbo green petrochemical industry cluster is not only expanding its output but is also transitioning towards lengthening supply chains and enhancing added value. The “14th Five-Year Plan” emphasizes focusing economic development on the real economy, while promoting smart, green, and integrated directions. For the Ningbo petrochemical sector, these three directions are equally relevant. This year, the Ningbo government proposed elevating the national-level green petrochemical manufacturing cluster to a world-class status.

In 2023, Ningbo identified the “361” industrial cluster as a core strategy for building a modern industrial system, which aims to cultivate three trillion-yuan industrial clusters including digital industry, green petrochemicals, and high-end equipment, alongside six hundred-billion-yuan clusters in emerging sectors such as new functional materials, new energy, and smart appliances.

Tu Jiong noted that benefiting from Ningbo’s robust industrial foundation and the characteristics of process chemicals, the city has consistently led the way in digital transformation across various industries, from intelligent manufacturing to 5G and industrial internet, as well as the future factory concepts and AI+ manufacturing.

Sinopec Zhenhai Refining has been a leading company in the Ningbo petrochemical industry, initiating digital transformation efforts early on. Chu Xiangping, Senior Expert on Information Technology at Zhenhai Refining, revealed that the company established a series of reform strategies, including deepening digital transformation in 2020. The company was also selected as one of China’s first 15 “leading smart factories,” being the only refining enterprise on the list. Zhenhai Refining began exploring digital delivery models in engineering as early as 2013 and has participated in formulating relevant national standards.

In terms of intelligent applications, a notable example is the wireless pump group model developed by Zhenhai Refining, which creates fault diagnosis maps based on pump speed, axial displacement, and vibration frequency parameters. By analyzing high-frequency and low-frequency energy components from vibration spectra, they have achieved comprehensive monitoring of multiple high-risk pumps across various units, thereby reducing manual inspection frequency by 50%.

In 2024, Zhenhai Refining collaborated with Control Technology to validate the TPT time-series large model, performing online fault prediction and diagnosis for nearly a hundred pumps in one facility. Over the course of a year, this model identified 11 issues across five typical fault types with a predictive accuracy of over 91%. Although the model requires ongoing training and optimization, it has established an effective risk warning mechanism capable of predicting equipment failures minutes or hours in advance.

Xue Baoxue, Chief Engineer at Wanhua Chemical (Ningbo) Co., Ltd., shared insights on the application of automation and intelligence in their quality control processes. By analyzing production data accumulated over time, they have applied intelligent methods to replace traditional manual inspections, significantly reducing sample analysis frequency and the workload of quality control staff. This not only enhances production efficiency but also ensures the safety and accuracy of product quality.

Wanhua Chemical has also partnered with Control Technology to explore the application of the TPT time-series large model in the process industry. At their caustic soda facility, they are examining intelligent control, analysis, and decision-making. Xue explained that applying AI directly in chemical production processes, especially those involving complex reactions, can be challenging. It is often better to combine mechanistic models of reactions with algorithmic models. Given that the chlor-alkali production industry is relatively mature and not bound by confidentiality concerns, it serves as an advantageous entry point for applying industrial large models.

Wanhua’s digital transformation efforts trace back to around 2014, initiated to improve quality and efficiency while enhancing competitiveness. The first implementation was a production safety alarm system. Xue emphasized that the future of intelligence, digitization, and AI should focus on production quality, efficiency, cost, reliability, and safety.

The chemical industry is inherently energy-intensive, and companies must continually seek technological means to reduce costs and enhance efficiency while persistently advancing digital and intelligent construction.

The application of AI large models in process-oriented industries presents significant challenges. Interviewees highlighted that safety is paramount, and the ultimate goal of technology is to reduce costs and improve efficiency. Zhao Lujun noted that process industry data exhibits distinct time-series characteristics, where parameter changes during production are influenced by historical data, process constraints, and equipment status. Traditional digital systems often rely on steady-state assumptions and only consider predictions under fixed conditions, while time-series large models can handle multiple variables and non-steady-state factors simultaneously for dynamic predictions and risk warnings.

These models capture complex causal relationships among key parameters such as flow, pressure, and catalyst activity in real-time, leading to more accurate predictions and optimal decision-making solutions. With the rise of new technological waves like ChatGPT, Control Technology recognized the profound impact of AI on the industrial sector and initiated a comprehensive “ALL in AI” strategy, unveiling the world’s first time-series large model, TPT (Time-series Pre-trained Transformer), in 2024.

For instance, the TPT model can provide risk warnings for process-oriented industrial production, ensuring safety while predicting equipment failures or abnormal process parameters based on dynamic data, thereby offering actionable warning solutions. Currently, TPT has been implemented across over a hundred industrial scenarios.

Zhejiang province is a major player in the petrochemical industry, with the province’s large-scale petrochemical chemical sector achieving an industrial value added of 1.7 trillion yuan last year, accounting for nearly 15% of the province’s industrial output. Presently, Zhejiang has established 14 future factories, 45 smart factories, and 118 digital workshops in the petrochemical sector. Dong Zhao, Deputy Director of the Zhejiang Economic and Information Technology Department, noted that while the digital transformation of the petrochemical industry is underway, challenges such as inaccurate supply-demand matching and weak capabilities among small and medium-sized enterprises persist, marking a critical phase of transitioning from isolated breakthroughs to comprehensive empowerment.

Qian Feng emphasized the importance of an industrial brain for AI applications in the petrochemical industry, stating that large models possess many advantages, but the key lies in how they integrate with specialized industrial fields, toolkits, and smaller models. The Ministry of Industry and Information Technology has organized the construction of a digital transformation scenario map and a four-list approach, focusing on data elements, knowledge models, software tools, and talent skills. This initiative aims to enhance the efficiency of supply-demand matching for transformation services, particularly in sectors like petrochemicals, steel, and machinery.

In the ongoing collection of best practices for digital transformation in the petrochemical industry, Zhejiang enterprises contribute nearly 60%. Ningbo continues to focus on unlocking the data value in its industrial sector. Tu Jiong mentioned that challenges remain in the digital transformation of the petrochemical industry, particularly regarding the release of industrial data and value, which is fundamental to AI. There are shortcomings in the aggregation, circulation, and application of industry data. To address this, Ningbo plans to develop a comprehensive toolchain for data collection, labeling, governance, and application, creating high-quality datasets for equipment maintenance, continuously improving the accuracy of predictive models, and conducting large-scale application pilots within the cluster.