April 2017

Process Engineering

Enhancements in EO/EG production technology

The markets for ethylene oxide (EO) and ethylene glycol (EG) continue to see attractive growth. This means that at least one or two new plants will need to be constructed every year if the industry is to meet forecast demand.

Van Milligen, H., Shell Global Solutions International BV; VanderWilp, B., Wells, G., CRI Catalyst Co.

The markets for ethylene oxide (EO) and ethylene glycol (EG) continue to see attractive growth. This means that at least one or two new plants will need to be constructed every year if the industry is to meet forecast demand.

To unlock capacity increases economically, new capital-efficient processes, higher-performance catalysts and continuous process optimization will be paramount. These aspects are considered here, beginning with a look at the innovations and technological achievements that brought the industry to where it is today.

Background and early history of EO catalysts

EO is produced by reacting ethylene with oxygen (O2) over a silver-based catalyst. These EO catalysts are characterized by several performance factors, including selectivity, activity, productivity and stability. One of the most important measures of an EO catalyst’s performance is its selectivity, which is the ratio of ethylene converted to EO to the total amount of ethylene reacted.

In the early days of EO production, the typical start-of-cycle selectivity for EO catalysts ranged from 68% to 70%—i.e., 30% or more of the ethylene feed to the process was lost to the complete combustion side reaction. Then, in 1971, an improvement was identified to the catalyst formulation that helped boost catalyst selectivity to greater than 80%.

Over the next 15 yr, the selectivity of EO catalysts seemingly plateaued, with only minor improvements to selectivity performance being realized. Some in the industry believed that EO catalyst advances had reached the theoretical limits of what was achievable. Since then, however, continuous improvements have been made to EO catalyst performance, including substantial increases in catalyst selectivity, stability and productivity.

Introduction of high-selectivity (HS) catalysts

In 1986, industrial catalyst researchers made a discovery that led to a significant improvement in process performance throughout the industry. As a result, one company was able to offer new catalysts to the market: HS catalysts. This discovery increased initial selectivity values by more than six percentage points to give start-of-cycle selectivity values of 86% or greater. The impact of this selectivity increase was huge, as it promised to save customers millions of dollars in ethylene feedstock costs.

Although HS catalysts provided great value with this selectivity boost, their activity and stability were lower compared with traditional catalysts, so they needed to be changed out more frequently. Traditional EO catalysts were changed out after 3 yr–4 yr of service; HS catalysts had a life of approximately 1.5 yr–2 yr.

Over time, both process and catalyst improvements extended the life and selectivity of HS catalysts. On the process side, new plants were designed with less severe operating conditions to cope with HS catalysts’ lower inherent activity. Notably, the new designs were based on lower volumetric production rates (referred to as the catalyst workrate) and lower reactor inlet carbon dioxide (CO2) concentrations. On the catalyst side, incremental advances in HS catalyst development led to initial selectivities of 90% or more, along with greater stability.

Consequently, by the turn of the century, new plants were benefitting from excellent selectivity and achieving a catalyst life of 3 yr or more. However, these benefits came with a compromise: increased capital costs. Larger reactors were necessary to compensate for the lower catalyst workrate, and larger CO2 removal units were installed to achieve low inlet CO2 concentrations, both of which resulted in increased CAPEX requirements. As the size of new world-scale EO production plants increased, the associated reactors became a significant capital (equipment) cost.

Introduction of high-performance (HP) catalysts

In 2010, yet another breakthrough in catalyst development provided an opportunity to address these capital costs. A global catalyst technology company developed a fundamental understanding of the catalyst aging process through focused research and development. This discovery led to new technologies that significantly inhibit catalyst aging, and resulted in the HP catalyst family.

HP catalysts are characterized by a high initial selectivity (comparable to that of HS catalysts), but with a significantly slower performance decline (Fig. 1). In addition, the HP catalyst family proved to be operable at significantly higher workrates, and to be tolerant of higher CO2 concentrations. These performance characteristics provided valuable opportunities for cost-saving changes to the process design.

Fig. 1. Example of selectivity curves comparing HS and HP catalysts over their lives at the same operating conditions.
Fig. 1. Example of selectivity curves comparing HS and HP catalysts over their lives at the same operating conditions.


EO/EG manufacturing process

Producing EO over a catalyst is the first step in the overall EO/EG manufacturing process. In the reaction section, EO is produced by catalyzed, direct partial oxidation of ethylene. Additionally, a portion of the ethylene fully oxidizes to form CO2 and water. These reactions take place in an isothermal (tubular) reactor at temperatures of 230°C–270°C. The reaction is moderated and optimized using an organic chloride. EO is recovered from the reactor product gas by absorption in water. Coproduced CO2 and water are removed and, after the addition of fresh ethylene and O2, the gas mixture is returned to the EO reactor as feed. The EO-water mixture can be routed to a purification section for recovery of high-purity EO and/or to a reaction section where EO and water are converted into glycols.

In the standard thermal glycol reaction process, EO and water are reacted at an elevated temperature (approximately 200°C) and pressure without catalyst. This process typically yields approximately 90%–92% monoethylene glycol (MEG) and 8%–10% heavier glycol products, mainly diethylene glycol (DEG) and triethylene glycol (TEG). The proportion of the higher glycols is limited by using excess water to minimize the reaction between the EO and glycols. The resulting water-glycol mixture from the reactor is then fed to multiple evaporators, where the excess water is recovered and largely recycled. Finally, the water-free glycol mixture is separated by distillation into MEG and higher glycols.

A more modern technology is to react EO with CO2 to form ethylene carbonate (EC) and subsequently react the EC with water to form MEG, with both reactions being catalyzed. In this two-step process, most of the MEG forms in an EO-free environment, which minimizes the coproduction of heavier glycols and results in an MEG yield of greater than 99%.

Fig. 2 shows a basic overview of the EO/EG process, which includes the following major sections:

  • EO reaction
  • EO recovery
  • CO2 removal
  • Light ends (LE) removal with optional EO purification
  • Glycol reaction and recovery
  • Glycol purification.
Fig. 2. Block scheme showing the major sections of the EO/EG process.
Fig. 2. Block scheme showing the major sections of the EO/EG process.

EO/EG process optimization

When designing an EO/EG plant, it is essential to take the owner’s priorities and business objectives into account. Experience shows that substantial value can be captured through process optimization. However, this involves highly complex evaluation, owing to the many parameters that influence project economics. It requires intimate knowledge of how specific operating parameters, such as feed gas composition and operating conditions (pressure, temperature, gas flowrate and catalyst workrate), influence key variables such as EO catalyst performance, capital investment and energy consumption. Examples of questions the designer must answer include:

  • How high should the gas flow be through the reactor? Increasing the flow enhances catalyst performance, but increases CAPEX and energy costs.
  • At what pressure should the EO reactor be operated? At higher pressure, the gas volume that is circulated through the reactor and in the recycle gas loop will be lower, which reduces energy consumption and equipment sizes. However, operating conditions become less favorable for catalyst selectivity, increasing feedstock consumption.
  • How low should the CO2 concentration be in the
    EO reactor feed gas? Reducing it will enhance catalyst performance, but this will also increase the capital and variable costs of the CO2-removal equipment system.

Table 1 shows how various operational parameters can have conflicting effects on feedstock use (related to catalyst selectivity), CAPEX and energy use. Table 1 shows, for example, that to minimize feedstock consumption, the catalyst workrate should be reduced, whereas minimizing CAPEX requires the catalyst workrate to be increased.


It is important to understand how these and other tradeoffs affect the overall cost of EO/EG production for a customer’s specific set of economic conditions. The process licensor should be able to help the owner select the design solution that offers the highest value, taking into account the owner’s specific circumstances and constraints.

One approach to arrive at the preferred design solution is to develop and apply proprietary models covering three aspects of the process:

  • A catalyst performance model that takes into account the EO reaction kinetics and calculates the catalyst performance profile for a defined set of process conditions
  • A CAPEX model that calculates investment costs
    for equipment and plant construction
  • An OPEX model that takes into account variable costs such as ethylene, O2, catalysts, chemicals and utilities,
    as well as the costs associated with changing out the EO catalyst, to arrive at the total variable cost of EO/EG production.

These models have been integrated to create a tool that helps plant designers predict the total EO production cost (Fig. 3) for a given catalyst and plant design. The data from the three models enable examination of the operating space to identify catalyst operating conditions that help minimize the total cost of producing EO. One learning from applying the tool is that optimizing the design conditions could easily save tens of US dollars per ton in EO manufacturing costs.

Fig. 3. The optimization tool leverages data from the three models and searches the  operating space for the best set of conditions for minimizing the total EO production cost.
Fig. 3. The optimization tool leverages data from the three models and searches the operating space for the best set of conditions for minimizing the total EO production cost.

Optimized process options

With both HS and HP catalysts available, optimized manufacturing processes can be developed and tailored to each catalyst family. In addition to the type of catalyst, the key differences between the two processes reside in the EO reaction system design. Compared with the HS process, the HP process enables 30% fewer reactor tubes, 20% shorter catalyst bed lengths, and significantly higher workrates.

Consequently, the reactor volume can be roughly halved. Depending on the overall plant capacity target, this reduction in volume may mean the elimination of one reactor from the lineup (Fig. 4). As the reactor(s) typically represent a significant portion of the capital required for new plants, the elimination of one reactor can substantially reduce capital investment costs.

Fig. 4. Since HP catalysts can operate at higher workrates,  a reactor can be eliminated from the lineup for certain capacity  ranges with the HP process.
Fig. 4. Since HP catalysts can operate at higher workrates, a reactor can be eliminated from the lineup for certain capacity ranges with the HP process.


Both processes provide low total costs for producing EO. The best choice of process option depends on relative CAPEX and raw material costs, which can vary geographically; and on considerations such as plant capacity and desired cycle time.

Debottlenecking existing plants

Opportunities are also available for existing producers. Many have significantly increased their output without major CAPEX, simply by upgrading to the latest-generation catalysts.

For example, Fig. 5 shows the catalyst selectivity performance over the life of three different generations of catalysts at one plant. In this example, the selectivity and life of the catalysts are shown (with each point representing approximately one month of operation). The first-generation HS catalyst gave an average lifetime selectivity of 84% during its 1.5 yr of operation. Performance significantly improved with use of the newer-generation HS catalyst, which gave an average lifetime selectivity of greater than 85% while operating at a higher workrate and for significantly longer.

Fig. 5. Real-world data showing how one EO producer benefitted by upgrading to latest-generation catalysts.
Fig. 5. Real-world data showing how one EO producer benefitted by upgrading to latest-generation catalysts.


The producer captured even greater benefits, however, when it upgraded to HP catalyst. With HP catalyst, the average selectivity is significantly higher, averaging approximately 87%, and catalyst life is much longer. All of these benefits have been achieved with the plant operating at record production rates—a workrate that is 14% higher than the average workrate of the previous-generation HS catalyst.

Key takeaways

The markets for EO and EG continue to see attractive growth, and the drive to improve EO/EG manufacturing economics is ongoing.

Advances in EO catalyst performance have been key factors in these improvements, providing higher workrate operations while improving selectivity and stability. EO/EG manufacturing processes have been optimized to take advantage of catalyst performance enhancements. These developments help to increase yields and reduce the CAPEX or running costs of grassroots projects. Existing plants can also benefit. Facilities around the world that have replaced older catalysts with latest-generation catalysts have captured substantial value.

Latest-generation processes that feature HP EO catalyst enable the EO reactors to be significantly smaller (up to 50%). As the reactor(s) typically represent a significant portion of capital required for new plants, the elimination of one reactor can substantially reduce capital investment costs. For some producers, especially those in areas where feedstock costs are high while capital is relatively inexpensive, leading-edge processes that feature HS EO catalysts may still be preferred.

Experience shows that substantial value can be captured by optimizing process design conditions. This optimization is complex because many design variables can influence the capital and operating costs in competing ways. A dedicated optimization tool has been developed to identify the preferred EO reactor operating conditions for each customer, taking into account local economics and constraints. HP

The Authors

Related Articles

From the Archive

Comments