The Industrial IoT Blog

If you’ve worked in Oil & Gas for a while, you know it’s a world apart — process industry, strict safety rules, and a patchwork of legacy systems that don’t like to talk to each other. Over the last two decades, I’ve led or architected SAP manufacturing solutions at a bunch of refineries, upstream and downstream operations, and gas processing plants across the Americas. If I had to sum it up, I’d say: nothing prepares you for the real grind of connecting SAP to a live refinery until you’ve stood in a control room at 2am, trying to explain to a shift supervisor why the new dashboard isn’t showing tank levels yet.

Getting Started: The Early Days with SAP xMII

I started with SAP xMII (what is now SAP MII) back in 2005. At that time, Oil & Gas plants were running on a mix of DCS (Distributed Control Systems), SCADA, and specialized historians. Most data lived in silos, and “integration” meant someone pulling numbers from a screen and typing them into SAP. My first refinery project was a crash course in process manufacturing — batch tracking, tank farm monitoring, lab results, and a constant stream of alarms from the control system.

One of my earliest big wins was architecting a solution that pulled real-time data from DCS and lab systems, gave operations a single dashboard, and let SAP see what was actually happening on the plant floor. We built reusable templates for batch genealogy, quality events, and shift handovers. That cut custom development time by half and set the stage for future accelerators and global standards. The project ended up winning a SAP Innovative Project Award, but honestly, the best part was seeing operators actually use the system to make faster decisions — not just for compliance, but for keeping the plant running safely and efficiently⁠⁠.

What Makes Oil & Gas Different

Here’s the thing: Oil & Gas isn’t like discrete manufacturing. You’re dealing with continuous processes, huge volumes, and a lot of regulatory pressure. Some of the unique headaches:

• Integration Complexity: DCS and SCADA systems are often decades old, with proprietary protocols. Getting SAP to talk to these systems takes a lot of custom connectors, middleware, and patience. Lab Information Management Systems (LIMS) add another layer — you need to pull in sample results, blend data, and batch records, all in near real-time.
• No Standard MES: Most sites had no coherent MES strategy. Some plants were highly integrated, others barely at all. Every site had its own custom apps for data collection, inventory, and production tracking. That meant every new SAP project started with weeks of reverse-engineering⁠⁠.
• Legacy Systems: It’s not rare to find a refinery running on a DCS from the 1980s, with data historians that only support basic trending. Bringing those into a modern SAP landscape means dealing with data quality issues, latency, and a lot of manual processes.
• Resistance to Change: Operators and engineers are used to their systems. If your SAP solution doesn’t make their life easier, they’ll find a way around it. I’ve seen more than one “integration” where the data was still being typed in by hand behind the scenes.

Technical Challenges and Real Solutions

DCS, SCADA, and Historian Integration

The hardest technical challenge was always integrating SAP with DCS and SCADA. You can’t just plug SAP into a refinery — you need middleware (like SAP PCo, custom OPC connectors) to collect time-series data, alarms, and events. Historian integration was another pain point: older historians didn’t support modern APIs, so we had to build custom batch retrieval scripts or use workarounds like CSV exports.

In one large refinery, we used SAP xMII to pull process data from a DCS, lab results from LIMS, and tank levels from a tank farm monitoring system. We built role-based dashboards for operators, maintenance, and management — each with their own KPIs. Getting the data mapped and normalized was a huge job. We ended up creating a Unified Namespace (UNS) model before it was even called that, just so everyone was working off the same tag names and units⁠.

Batch Tracking and Tank Farm Monitoring

Batch tracking in Oil & Gas is about genealogy — knowing every input, every process step, and every test result. We built templates in SAP MII that tied together process orders from SAP ERP, batch events from the DCS, and lab results. For tank farm monitoring, we had to integrate real-time tank level sensors, temperature, and movement data. The solution was a set of custom connectors and dashboards that let operators see tank status and batch movement live, with alerts for deviations or safety issues. This was not plug-and-play — every site had its own quirks, and we spent a lot of time mapping data points and validating with operations⁠⁠.

Templates, Standards, and Change Management

One thing I learned the hard way: standardization saves you in the long run. Early projects were all custom code. Later, we developed reusable templates for common functions — batch tracking, shift handover, quality event dashboards, and asset monitoring. This not only sped up delivery but made support and upgrades much easier.

Change management was (and still is) a constant battle. You have to work with operations, automation, and IT to define architecture, configure integrations, and train users. If you skip user training or don’t involve operators early, the project will stall. I always tried to spend time in the control room, listening to what folks actually needed, not just what IT thought was required⁠.

Lessons Learned from Real Projects

• Start Small, Prove Value: Don’t try to boil the ocean. Start with one use case (like batch tracking or tank monitoring), show ROI, then expand. Early deployments are the toughest — capture lessons learned and use them to build templates for future rollouts⁠⁠.
• Design the Data Model Upfront: A poorly designed data model (UNS, tag naming, units) will haunt you forever. Invest the time to get it right before you start building dashboards or analytics⁠⁠.
• Stakeholder Engagement: Engage IT, OT, and business users from the start. If you don’t have buy-in from the folks running the plant, you’ll end up with shelfware.
• Integration is Never “One and Done”: Every site is different. You need flexible, configurable connectors and a way to handle legacy systems. Sometimes you have to use creative workarounds to bridge old and new tech.
• Security and Compliance: Oil & Gas is a high-risk environment. Every integration needs to be validated for cybersecurity, safety, and regulatory compliance. We adopted segmented network zones, role-based access, and strict change control to meet internal and external requirements⁠.

What Worked (And What Didn’t)

What worked best was focusing on real-time visibility — giving operations a single version of the truth, from the shop floor to the top floor. SAP MII let us build dashboards that pulled together data from DCS, historians, lab systems, and SAP ERP. When done right, this enabled faster, better decisions and moved plants from reactive to proactive operations⁠⁠.

What didn’t work was over-engineering. I’ve seen projects fail because they tried to automate everything or forced operators to change workflows overnight. Sometimes, a simple dashboard and a CSV export do more good than a fancy analytics platform nobody uses.

Honest Opinion

I’ll be honest: SAP MII was a game-changer for Oil & Gas, but it was never easy. The tech is only half the battle — the real challenge is people, process, and legacy culture. If you want a project to succeed, spend as much time on the shop floor as you do in design workshops. And don’t be afraid to admit when something isn’t working — sometimes the right answer is to step back and keep it simple.