By Cody Berra, CMRP, Senior Solutions Consultant, UptimeAI
Three pressures, one operational question — and why agentic operations is becoming a C-suite conversation.
Ten years ago, the nitrogen playbook was relatively straightforward. Success came down to driving cost down, keeping rates high, and executing turnarounds cleanly. That model held up in a more stable market, where variability existed but could largely be managed within known bounds.
After three days at CRU Nitrogen and Syngas 2026 in Dallas, it is clear the above operating model is no longer sufficient for what is coming next. The kickoff session from Justin Rackleff, Principal Analyst, Fertilizer at CRU put a fine point on the shift. Roughly one third of global fertilizer supply has been disrupted by geopolitical conflict, forcing trade flows to be rerouted in real time rather than optimized over longer horizons. Prices are reacting quickly to these supply shocks, and even producers in the United States, despite their structural cost advantage, are feeling the impact of a more interconnected and volatile global market.
At the same time, the expectations placed on producers are increasing. There is a push to grow nitrogen and syngas output while also lowering carbon intensity, which introduces a layer of operational complexity that cannot be solved through capital investment alone. These pressures are not temporary dislocations that will normalize in a few quarters. Volatility, decarbonization, and the ongoing loss of experienced operators and engineers are becoming defining characteristics of how plants are run.
Taken together, this shifts the conversation from strategy to execution. It is no longer just about where you sit on the cost curve, but how effectively you can respond as conditions change around you. That reality ultimately comes down to a single operational question, and the producers who can answer it consistently will separate themselves over the next decade.How quickly can your plant turn an alarm into the right action?
The Cheapest-Producer Thesis Is Breaking
For a long time, the answer to who wins in nitrogen production was simple. The lowest cost producer had the advantage, and over time that advantage compounded. In a more stable environment, that logic held up.
What is changing is not that cost no longer matters, but that it is no longer enough on its own. In a market defined by volatility and tighter operating constraints, the ability to make the right decision at the right time is starting to matter just as much as the underlying cost position. Decision velocity, more than cents per ton, is beginning to shape outcomes.
That showed up repeatedly in conversations throughout the week. The same underlying issue, something like early-stage fouling or a subtle process imbalance, can play out very differently depending on when and how it is addressed. When it is identified early, it becomes a manageable event that can be worked into a planned intervention. When it is missed or misdiagnosed, it has a way of escalating into an unplanned outage with meaningful financial impact.
The difference is not the asset or even the condition itself. It is the speed and quality of the response. In a more volatile market, that gap becomes more expensive, because there is less room to absorb inefficiency or downtime. Small delays in recognizing what is happening, or uncertainty in what to do next, quickly translate into lost margin.
This is where the traditional cost curve starts to break down. Cost will always be a factor, but it does not protect against slow or incorrect decisions. The producers who are pulling ahead are the ones who can consistently turn plant signals into the right operational action faster than the market can penalize a mistake. Over time, that capability compounds in the same way cost leadership once did.
Decarbonization is now an operations problem, not just a capital one
The decarbonization conversation in nitrogen has historically been framed as a capital problem. Most of the focus has gone toward large investments like CCUS, blue ammonia, or electrolyzer-driven hydrogen. Those initiatives still matter and will continue to shape the long term direction of the industry.
What came through more clearly this week is that carbon intensity is also a function of how well plants are run today. It is not just about what gets built next, but how close existing units operate to their intended design.
A reformer running a few percentage points off target is not only leaving margin on the table, it is also increasing the carbon intensity of every ton produced. The same is true for fouling that goes undetected, heat transfer inefficiencies that persist longer than they should, or operating conditions that drift without being corrected. These are not just reliability or performance issues anymore, they are directly tied to emissions.
Decision velocity is now an ESG variable, highlighting the need for scaling expertise.
In practice, that means decarbonization is no longer separate from day to day operations. Every delayed cleaning, every unplanned shutdown, and every period of suboptimal operation carries a carbon consequence that has to be accounted for elsewhere. Over time, that adds up in ways that are difficult to offset through capital projects alone.
For producers investing in blue ammonia or low carbon hydrogen, this becomes even more important. Those projects are often modeled assuming a stable and well understood operating baseline. When the underlying unit is not consistently running close to design intent, that baseline becomes less reliable and the economics become harder to defend.
The expertise crisis is already in the math
The workforce challenge came up in a more direct way than in prior years. It is no longer a future concern, it is already affecting day to day operations. When an experienced operator or specialist leaves, the impact goes beyond headcount. It is the loss of unit specific judgment built over decades, including how to interpret weak signals, recognize early deviations, and respond under non ideal conditions. That kind of knowledge is difficult to document and even harder to transfer in a way that is immediately useful.
At the same time, the next generation is stepping into a more complex environment. Plants are dealing with more variability, tighter carbon constraints, and new technologies layered onto existing assets, all with less institutional knowledge to rely on.
This is where the earlier pressures start to converge. Volatility increases the cost of slow decisions, decarbonization raises the stakes on how precisely plants are run, and the loss of experience reduces the amount of expert judgment available at any given moment.
That gap is difficult to close through hiring alone. The constraint is no longer access to data, it is access to consistent, scalable decision making. And this decision-velocity gap, can only be resolved by encoding expertise into the system itself.
Agentic operations is what closes the gap
Taken together, these pressures point to a shift in how plants need to operate. The question is no longer whether more data or better visibility is needed. Most sites already have that. The challenge is turning that information into the right decision quickly and consistently.
This is where the conversation around agentic operations is starting to take shape. Not as autonomous control, and not as another layer of dashboards, but as a way to close the gap between detecting an issue and knowing what to do about it.
At most sites today, the real bottleneck is expert decision capacity. The ability to interpret signals in context, connect them to likely failure modes, and determine the right course of action is still heavily dependent on a small number of experienced individuals. As those individuals become harder to scale, the need to support that decision process in a more systematic way becomes more obvious.
That is the role these systems are beginning to play. They are not replacing operators or engineers, but helping extend their judgment across more assets, more conditions, and more moments in time than would otherwise be possible.
Three major shifts where agentic operations can empower senior leaders
- From cost optimization to decision optimization. Cost will always matter. In this market, the producers who pull ahead will be the ones turning signals into action faster than their peers — at every site, every shift, every loop.
- From workforce planning to expertise institutionalization. You cannot hire your way out of the retirement math. You can encode the judgment of your best operators and engineers into systems that operate at machine speed across the fleet.
- From CCUS-as-investment to operations-as-decarbonization-lever. Carbon intensity now lives in how well today’s plants run, not just in tomorrow’s projects. The producers treating operational excellence as ESG strategy will compound an advantage the producers treating it as cost will miss.
We do not know what next year’s CRU will bring, but the conversation is already moving from “is this real?” to “how fast can we deploy?” That is a healthy shift for an industry that historically does not move first. Contact us to uncover how UptimeAI reasoning agents turn expertise into real-time, scalable advantage.
