Update on Sulfuric Acid Dew Point Monitoring – Applications and Case Study

Background

In the Fall/Winter 2021 Edition of Sulfuric Acid Today, Dan Menitti with Mississippi Lime Company (now MLC) authored an introductory article (1) that detailed the development and introduction of the first viable “dew point” monitor to be applied to the sulfuric acid manufacturing industry.

That instrument, known to many as the Breen SA probe, has since logged many additional years of uninterrupted operation in multiple acid plants throughout the US. The performance of those instruments in providing early detection of moisture leaks is well documented and has been corroborated by those installations many times since then.

In 2023, the technology behind the Breen ABS probe and the Breen SA probe was acquired by Ohio Lumex, a company located near Cleveland, Ohio. While the basic operating principles behind the technology were kept unchanged, many design improvements to the mechanical and electrical systems were implemented during 2023 and 2024. In late 2024, the Ohio Lumex Ei4200 Dew Point Monitor was released to multiple industries.

The intent of this article is to present some of those updates and to document the performance of the Ei4200 compared with its predecessor. Additionally, data from the Ei4200 running in active mode at additional US sulfuric acid plants is presented for consideration.

Mechanical Design Upgrades

The Ohio Lumex mechanical redesign addressed several long-standing issues with the Breen system. These included:

• Change in Heating System – Both the Breen SA and the Ohio Lumex Ei4200 include a supplemental heating system to provide sufficient heat at the sensor to evaporate any residual, high-boiler-temperature, compound. Where the Breen system employed a contact system surrounding the sensor cell, Ohio Lumex employs an in-line heater that uses the “cooling air” to heat the cell. This allowed the Ei4200 to eliminate the shrouding around the sensor, facilitating on-site cell maintenance when necessary
• Field Replaceable Cell – As the cell is a glass component it is the most fragile point in the process and does, from time to time, need maintenance or replacement. By changing the mechanical design of the cell housing, the Ohio Lumex approach allows the cell to be removed, cleaned and/or replaced in the field by plant I&C staff. The process is simple requiring only four (4) bolts and three (3) wires to be addressed. Total time to replace a cell in the field should be less than one hour
• Leak Check Detection – Movement of the process gas isolation valves to a location internal to the probe control box allowed a reduction in physical probe size and allows the system to be pressurized and isolated on a programmable basis to check for cell integrity and/or leakage

Electrical Design Upgrades

The key change to the electrical system focused on a dramatic improvement in current detection sensitivity.

Operating Approach

The Ei4200, since it uses the same basic sensory cell as the Breen SA, employs the same base operating approach:

1. The temperature of the cell is verified to be above the “start temperature”
2. Cooling air enters the cell chamber from the backside, through a small, insulated tube
3. Cooling air flow is controlled by a precision flow meter, providing accurate temperature reduction of the sensor tip at a configurable rate
4. When the temperature of the sensor drops below the condensation temperature of the gas, acid vapor will begin to condense on the cell surface
5. Embedded electrodes in the sensor surface detect both the temperature of the sensor and the level of current flow on the sensor surface
6. When the sensor surface current exceeds a configurable delta limit, the presence of condensation is noted, and the temperature at which the current delta limit is exceeded is recorded.
7. At this point, the cooling air flow to the cell is restricted in a configurable fashion and the temperature of the sensor increases until it reaches start temperature and the process starts again.

This entire process is depicted in the bottom graphic in Figure 2 with a close-up of the process in Figure 3 below:

There are multiple operating modes that employ variations of the above process:

• Active Mode – In active mode, the cycle described above operates continuously with one cycle immediately following the previous one. This method provides the most up-to-date information on flue gas condensation temperatures (dew point) and the most sensitivity to changes that represent potential increases in system moisture leakage.
• Passive Mode – In passive mode, the system remains dormant at the nominal flue gas temperature, but the detection circuitry remains active. If a moisture leak occurs that pushes the condensation temperature up to the flue gas temperature, the system will detect this change and report it to the DCS. On a programmable basis, the instrument will execute an “Operability Check” to provide regular readings of flue gas condensation temperature and to verify the system remains operational.
• Hybrid Mode – This mode is like the legacy Breen mode in that it operates a continuous cycle, but the start and minimum temperature limits are configured marginally above the gas condensation point. This provides the highest sensitivity to changes in flue gas moisture, while also reducing compressed air consumption and minimizing corrosion of the sensor electrodes. However, it does not provide a continuous condensation temperature reading and, therefore, cannot provide continuous input to other potential process changes.

Dew Point vs. Condensation Temperature

Historically, this instrument, along with its predecessors from Breen and earlier from Land Instrument, looked for an “equilibrium dew point”. This definition of dew point can be found in many research papers, but essentially, it is the temperature at which the gas achieves 100% saturation of the vapor in question. With this definition in mind, the Breen Instrument referred to two gas parameters. The first, “dew point”, represented equilibrium between vapor condensing from the gas onto the cell and liquid evaporating from the cell into the gas. The second, “formation temperature,” represented the temperature where some level of condensation was first detected on the cell.

Multiple field trials, including Breen probes and various wet chemistry analytical techniques, provided support for the “equilibrium dew point” approach as the detected concentrations fit well into widely accepted formulas linking SO₃ concentration, H₂O concentration, and projected dew point. Breen primarily utilized the Verhoff-Banchero formulas for the calculation of SO₃ from measured dew point.

However, more recent work documented by Li, et.al, ⁽²⁾ in 2019 and follow-on work by Zuo et. al, ⁽³⁾ in 2020 suggest that better instrumentation methods available today show that the true “dew point” of vapor in a gas is the temperature at which that gas just begins to condense.

Both papers reference a “state diagram” ^(2,3) that supports the notion that historical methods of predicting dew point utilized the concentration/temperature associated with the highest rate of condensation rather than the initial temperature.

Operating Data

The Ei4200 now boasts multiple installations at sulfuric acid plants, industrial dry scrubber operations, and monitoring of acid gas concentrations post-dry sorbent injection operations at coal-fired power plants. For this discussion, data from two separate sulfuric acid plants will be presented and discussed. Both are sulfur burners. Both plants have the instrument installed after the secondary economizer and ahead of the absorption tower.

Installation “A”

Figure 5 shows the operation of plant “A”, and the response of the Ei4200 to changes in plant operations.

There are several points worth highlighting in this data set:

1. The blue line represents the delta current limit that will trigger the identification of condensate. The unit had been operating at higher current levels to start and was slowly being tuned to find the lowest stable Delta current settings.
2. Note the long-term shift in “dew point” as the set point was lowered from 0.2 uA to 0.1 uA. The increase is to be expected since the trigger is happening with less condensate on the sensor and thus earlier in the temperature descent cycle. The dew point differential is in the 10°F – 15°F range.
3. High current events occurred three times during this period. 9/25, 10/1 and 10/4. While the operating conditions around each event are somewhat different, the result on the instrument and the instrument’s response in all cases were the same.
4. Figure 6 shows the same data, but with a high-resolution look at 10/1 for specifics:

Of note from the High-Resolution Data:

• The expected and repeatable pattern between changes in delta limit and reported dew point.
• There is some overlap in trace in the late morning, so Figure 7 below shows the current waveform (with the green color highlighted) and without the associated cycle temperatures
• When the delta limit was increased just after 9:00 AM, the current traces climbed, and the dew point dropped as you would expect.
• However, when the delta limit was dropped around 1:00 PM, two short cycles followed with a major current spike after that. This reflects an error in setting optimization. The start temperature for any subsequent cycle may be automatically adjusted after each cycle to minimize the total cycle time based on the actual detected dew point. However, in this case, the calculated Start temperature was too low for the conditions, and the current from the preceding cycle had not developed. This led to a large amount of condensate forming on the sensor tip before it cleared itself based on the longer-than-usual and higher-than-usual cycle temperature peak (this can be noted by referring to the previous figure)
• Perhaps more importantly, when the delta limit was increased to help offset this issue, four strange current wave forms appeared before the system stabilized, on its own, and returned to normal operation.
  • These four wave forms reflect an interaction between the high current (liquid) deposit on the tip and gaseous compounds, likely SO₃. In practice, as the cycle temperature rises, the liquid begins to evaporate, but before it can completely clear a second compound manifests itself and requires a higher temperature to evaporate.
  • As can be seen, with each succeeding cycle, there is less secondary high temperature required to eliminate the compound, and eventually, there is not enough residual liquid on the tip to interact with the SO₃. At this point, the disturbance cleared.

This level of detail is presented because the same phenomenon appeared to drive each of the disturbances from 9/25. 10/1 and 10/4. While the software is being updated to automatically recognize and compensate for this known phenomenon, what caused it to happen on 9/25 and 10/4 when the auto start feature was off appears to be plant process-related.

Is there a value to plant operators to know when there is a difference in apparent process gas makeup? Only more experience will tell.

Installation “B”

A different Ei4200 was placed into service at a plant on the East Coast on April 1, 2025. The data presented below will contain less narrative, as it is hoped that the reader has already gained some understanding of the unit process and constraints.

For Example B, only the reported dew point (condensation point) has been reported over a period from June 22nd through July 14^th.

There are two points for discussion here: the tall yellow highlighted ovals centered roughly around July 10^th, and then a series of smaller, orange ovals that appear at semi-regular intervals throughout the operating period.

1. The plant reported that a series of major thunderstorms passed through the area over that period of time. It is explained by the plant that the amount of rain that fell apparently overwhelmed the combustion air dryer, resulting in a higher level of moisture entering the process.
2. A rise in general-trend dew point is associated with three separate time periods.
   1. 6/22 – 6/276/29 – 7/7
   1. 7/8 – 7/14
3. During these three separate time frames, the delta current was being adjusted to better understand the actual temperature where condensation first appeared. This has been explained in the prior example.
4. What jumps out in this example is, again, the orange ovals. In each of these ovals, a rapid drop in dew point is reported following a gradual increase. While a general fluctuation in dew point is to be expected, what is not yet explained is why these dramatic drops in dew point occur and what meaning/value can it provide to assist plant operations in improving yield and/or reliability.

Conclusions:

At this point in time Ohio Lumex is quite pleased with the demonstrated performance of the re-engineered Breen system that is now the Ei4200.

Some stated observations:

• While none of our instruments has, as yet, experienced a cell breakage during operation, one cell did break during transit. This gave us an opportunity to work with the plant staff to replace the cell on-site. We are pleased it took less than an hour. When the cell was pressurized (leak test process) a leak was found. Further inspection found that one of the four mounting bolts had not been tightened. When both were tightened, the leak test passed, and That instrument has been in virtually consistent operation for about 4 months.
• The current measurement process has been wildly successful. Measuring deposition-based currents with granular readings of 0.01 uA allows significantly faster cycles with significantly less material buildup and long-term corrosion.
• However, being able to read information at sub-micro amp levels has revealed some liquid-gas interactions that were not apparent with the much higher reading Breen system
• It is currently projected that an Ei4200 will operate in active mode within the duct of a sulfuric acid plant for at least 12 months. At that point, a simple cell replacement should offer another 12 months of operation with maintenance costs limited to onsite plant staff
• We are anxious to continue working with forward-looking operations to identify the basis for the observed changes in Ei4200 readings and identify potential changes to sulfuric acid plant operation that can be of long-term benefit.

We wish to offer our sincere thanks and appreciation to all those who are currently using, trialing, and/or exploring the use of this instrument for their operations. The information that we gain from each and every experience helps us expand the reach and value of this technology.

References:

• “The latest in sulfuric acid plant process gas dew point/moisture leak detection”; Daniel T. Menniti, Director of Business Development; Mississippi Lime Company/Breen, 3870 S Lindbergh Blvd, St. Louis, MO 63127
• “Experimental method for observing the fate of SO₃/H₂SO₄ in a temperature decreasing flue gas flow: Creation of State Diagram”; Yuzhong Li^(a), Qingwu Zhu^(a), Qiujie Yi^(b), Wujun Zuo^(a), Yupeng Fenga^(c), Shouyan Chena^(d), Yong Dong^(a); (a) National Engineering Laboratory of Coal-fired Pollutants Emission Reduction, School of Energy and Power Engineering, Shandong University, 17923 Jingshi Road, Jinan 250061, China; (b) College of Mechanical and Electronic Engineering, Shandong University of Science and Technology, 579 Qianwangang Road, Qingdao 266590, China; (c) Shandong Low Carbon Expert Sci. & Tech. Co. Ltd., 54 Maanshan Road, Jinan 250002, China; (d) Shared Laboratory of Energy and Environment, Shandong University Science Park, 54 Maanshan Road, Jinan 250002, China
• “Review of flue gas acid dew point and related low temperature corrosion”; Wujun Zuo, Xiaoyu Zhang, Yuzhong Li; National Engineering Laboratory for Reducing Emissions from Coal Combustion, Engineering Research Center of Environmental Thermal Technology of Ministry of Education, Shandong Key Laboratory of Energy Carbon Reduction and Resource Utilization, School of Energy and Power Engineering, Shandong University, Jinan, Shandong, 250061, China