October 2021

Hydrogen

Monitoring hydrogen plant performance—Part 1

Process monitoring is an indispensable practice to keep track of key performance indicators (KPIs) of the hydrogen (H₂) plant.

Process monitoring is an indispensable practice to keep track of key performance indicators (KPIs) of the hydrogen (H₂) plant. A good system of process monitoring not only ensures safe and reliable plant operations, but also helps in making strategic decisions such as catalyst changeout schedules. If KPIs are not monitored closely, there can be situations where the expected yields are not achieved. This affects the economics of the H₂ plant and that of the entire refinery complex, since the downstream hydrotreated/hydrocracked product yields are affected, as well. This will have a direct implication on the overall refinery margin.

For better monitoring and control, it is imperative to know the fundamentals of performance indicators. Performance monitoring requires operating data inputs from the plant, which are normally accessible from the intranet servers or from the distributed control system. It also requires stream analyses from the laboratory (or from online analyzers) to understand how each reactor is performing with respect to conversion/yield. The laboratory testing frequency of such streams is normally decided by the process engineering department in conjunction with the laboratory section.

The main objective of this article is to guide H₂ plant process engineers in:

• Performing a detailed mass balance across the H₂ flowsheet by using available information, such as dry analysis of outlet streams. Doing so will help identify bottlenecks across each reactor. The focus would be to see how much H₂ is being made before the final stream enters the pressure swing adsorption (PSA) unit. This mass balance will also help estimate the outlet stream’s composition on a wet basis, thereby facilitating the estimation of equilibrium constants (K_eq values) for steam methane reforming (SMR) and water gas shift (WGS) reactions, which will help calculate the approach to equilibrium (ATE) values. These values, which are important in understanding the catalyst activity, can then be compared with the kinetic model values provided by the catalyst supplier.
• Monitoring other critical parameters and KPIs across each reactor in the H₂ flow sheet.

Feed definition and characterization

Two types of feed are considered for studying plant performance: Case 1 is a natural gas (NG) feed, and Case 2 is a naphtha feed. Note: The flow, composition and operating conditions are purely assumptions based on the author’s experience and do not correspond to any specific plant.

For Case 1, feed comprising 1,000 kmol/hr of (NG + recycled H₂) is assumed to have the following composition:

• Hydrogen (H₂) = 2%
• Nitrogen (N₂) = 5%
• Methane (CH₄) = 90%
• Ethane (C₂H₆) = 2%
• Propane (C₃H₈) = 1%.

This combined stream of NG and recycled H₂ contains 10 ppmv of hydrogen sulfide (H₂S) and 5 ppmv each of methyl mercaptan and dimethyl disulfide (DMDS).

For Case 2, 80 kmol/hr of naphtha, with 20 kmol/hr of recycle H₂, is assumed. The most common laboratory information for naphtha is the ASTM D86 standard covering distillation and specific gravity. The ASTM D86 cut points of the assumed naphtha feed are:

• T_10% = 110°F
• T_30% = 128°F
• T_50% = 135°F
• T_70% = 145°F
• T_90% = 155°F.

The reported specific gravity is 0.716. Naphtha must be characterized further based on the above information. Degrees Fahrenheit and Kelvin will be used only for the following estimations.

Naphtha characterization. Eq. 1 is used to find the volume average boiling point (VABP):

(T_10% + T_30% + T_50% + T_70% + T_90%)/ 5 = 134.6°F                                                                                                                       (1)

The correction factor is:

• Slope = (T_90% – T_10%)/ 80 = 0.563
• Ln ∆ = [–0.94402 – 0.00865 (VABP – 32)^0.6667 + (2.99791 x slope^0.333)]
• ∆ = 3.825°F.

The mean average boiling point (MeABP) is shown in Eq. 2:

MeABP = VABP – Δ = 134.6 – 3.825 = 
   130.8°F = 328.03°K                                                                                                                                                                     (2)

The molecular weight is calculated using Eq. 3:

1.6607 x 10^–4 × (MeABP)^2.1962 ×
   (specific gravity)^–1.0164 = 1.6607 × 10^–4 ×
   (328.03)^2.1962 × (0.716)^–1.0164 = 78.2 kg/kmol                                                                                                                                 (3)

The carbon/hydrogen (C/H) weight ratio is calculated using Eq. 4:

3.4707 [exp ^{{(0.01485 Tb + 16.94x (specific gravity) – 0.012492 Tb × (specific gravity)}} × T_b^–2.725 × (specific gravity)^–6.798]                                              (4)

Substituting T_b = MeABP; therefore, the C/H weight is 6.022.

The molar H/C = 12.01/(C/H weight), which equals 1.9944. The empirical formula of naphtha is CH_1.9944. The empirical molecular weight is shown in Eq. 5:

1 × 12 + 1.9944 × 1.008 = 14                                                                                                                  (5)

The carbon number of naphtha is found by using Eq. 6:

Actual molecular weight/empirical molecular weight = 78.2/14 = 5.586                                                         (6)

Feed purification, Section 1: Hydrogenator

The feed, along with recycled H₂, first enters the hydrogenator reactor. One of the primary functions of the hydrogenator is to convert organic sulfur (S) compounds (e.g., mercaptans, sulfides, disulfides and thiophenes) to H₂S. Other primary functions are to convert organic chlorides (if present) to hydrochloric acid (HCl) by reacting with H₂ and to hydrogenate olefins that may be present in the feed.

Key parameters to be monitored include:

• H₂ level for each type of feed or combination of feeds. This level is generally suggested by the catalyst supplier. The process engineer should confirm this level with the catalyst supplier in case of any deviation in feed specifications. The recommended H₂ level in the outlet stream is normally 2% for NG feeds and about 26% for highly aromatic naphtha feeds.
• The optimum operating temperature is 350°C–400°C (662°F–752°F). For the LPG feed, the maximum temperature limit may be lower due to susceptibility to form carbon. Operating higher than the recommended maximum temperature may cause carbon deposition on the catalyst and upstream pre-heat coil, leading to high pressure drop issues. Operating lower than the recommended minimum temperature may cause organic sulfur or chloride to slip through and poison the downstream reforming catalysts.

Among the KPIs is pressure drop. Being the first reactor in a typical H₂ flow sheet, the hydrogenator is vulnerable to pressure drop issues. Some general reasons for increases in pressure drop across any fixed bed include:

• Breakage or erosion of catalyst particles, primarily due to poor handling and loading, is one of the causes of high pressure drop.
• Disintegration of catalyst pellets—primarily the top layer—is another cause of pressure drop due to poor inlet gas distribution and/or an inadequate hold-down layer on the top. In some cases, the disintegration of poor-quality support balls in the hold-down layer contributes to high pressure drop.
• Carryover on the catalyst bed is one of the most likely reasons for high pressure drop across the hydrogenator. Any debris upstream of the reactor, if not properly removed, can get carried over to the top of the catalyst, leading to a high pressure drop. Specialized foulant trapping materials are commercially available to address this issue.
• Deformation of catalyst pellets due to accidental wetting of the catalyst, causing a decrease in catalyst strength, can lead to deformation and, in the worst case, disintegration. This issue can be encountered in high-temperature or medium-temperature shift reactors where there is a likelihood of upstream boiler water leaks.
• The collapse of the bed support grid or any damage to the outlet collector can result in a significant pressure drop.
• Operating the reactor (hydrogenator and pre-reformer) at higher than recommended temperatures can cause thermal cracking of the hydrocarbon feed, thereby depositing carbon over the catalyst, leading to a high pressure drop. In addition, for pre-reformers processing naphtha feeds, there is a minimum bed temperature below, which could have issues of polymeric carbon formation that can lead to pressure drop increases.

It is important to monitor the pressure drop trend closely from the time the catalyst has been put in operation. If the pressure drop increases suddenly, then that time (before and after) should be isolated and investigated in detail. This generally applies to all reactors in the flowsheet.

Inlet and outlet chlorides and sulfur. It is important to monitor the inlet and outlet levels of chlorides and sulfur. In most H₂ plants, the feed is analyzed daily, so the total inlet chlorides and sulfur are known. The outlet H₂S and HCl can be routinely measured by process engineers using detector tubes, or they can be analyzed in a laboratory. The outlet H₂S measurement will help determine if the hydrogenator is converting all organic sulfur to H₂S.

For example, if the feed (NG + recycle H₂) in Case 1—containing 10 ppmv of H₂S and 5 ppmv each of methyl mercaptan and DMDS—is passing through the hydrogenator, then the outlet should measure the calculated H₂S value in Eq. 7:

10 ppmv H₂S + 5 ppmv × 1 (methyl mercaptan) + 5 ppmv × 2 (DMDS) = 25 ppmv of H₂S                                                                         (7)

Feed purification, Section 2: Chloride absorber

This section is not common and is only required when the feed contains chlorides. The primary function of the chloride absorber is the absorption of HCl, which is either present in the feed or formed across the hydrogenator. In most cases, this catalyst is installed as a small layer below the hydrogenator. In some flowsheets, it is also installed above the H₂S absorber. KPIs include the following:

• Outlet chloride should be less than 0.1 ppmv vs. inlet chloride.
• The catalyst life depends on pickup/capacity. The process engineer should have the information of expected pickup from the catalyst supplier.

Feed purification, Section 3: H₂S absorbers (zinc oxide)

The primary function of the H₂S absorbers is to absorb H₂S, which is present either in the feed or formed across the hydrogenator. This section is usually seen as two vessels in a lead-lag arrangement or as a single vessel when sulfur levels are very low. In addition to pressure drop, KPIs include the following:

• Outlet H₂S should be less than 0.1 ppmv vs. inlet H₂S.
• The catalyst life depends on pickup/capacity. The process engineer should have the information of expected pickup from the catalyst supplier.

Note: The process engineer should ask the catalyst supplier for actual pickup and not theoretical pickup. The theoretical pickup might be much higher than the actual pickup.

For example, Case 1—1,000 kmol/hr of (NG + recycle H₂) with 25 ppmv of H₂S—would have a theoretical pickup of 600 kgS/m³ and an actual pickup of 500 kgS/m³. To reiterate, the pickup figures are an assumption and do not correspond to any specific product.

The volume of each bed is 10 m³. Eq. 8 is used to calculate the amount of sulfur to be removed:

(25/1,000,000) × 1,000 kmol/hr × 24 = 0.6 kmol H₂S/day = 20.4 kg H₂S/day = 20.4 × 32/34 = 19.2 kgS/day                                                            (8)

Therefore, the expected life is calculated using Eq. 9:

500 kgS/m³ × 10 m³/19.2 = 260 days                                                                                                                                                                     (9)

The most economical way of operating a lead-lag arrangement is to allow the lead bed to operate until near saturation—i.e., until the outlet H₂S from the lead bed is 90% of the inlet H₂S.

Feed purification, Section 3: Ultra-purification

This section is not common in all flowsheets. The primary function of this section is to purify H₂S to parts per billion (ppb) levels. It is usually seen as a small layer beneath the zinc oxide absorber(s). KPIs include the following considerations:

• It may be difficult to measure very low ppb levels of H₂S by using detector tubes or the laboratory. However, the economic benefit of having this layer can be realized in terms of extended life of the pre-reformer catalyst. Ultra-purification also helps extend the life of primary reformer catalysts in highly stressed reformers.
• H₂S pickup/capacity is another KPI.

Pre-reformers

Pre-reformers are not found in all flowsheets. The most important benefit of a pre-reformer is the flexibility to process different feeds. The primary function of the pre-reformer is to convert C₂+ in the feed to C₁ and H₂. The key parameters include the steam-to-carbon (S/C) ratio, inlet temperature, pressure, Z₉₀ progression and pressure drop. Z₉₀ is an important trend that needs to be monitored closely, since it determines the timing of the pre-reformer catalyst changeout. The typical Z₉₀ plots for NG and naphtha feeds are shown in FIGS. 1 and 2.

KPIs include C₂+ slip, pressure drop and Z₉₀ progression. ATE is not a KPI for the pre-reformer, as the SMR’s ATE would be close to zero throughout the expected life of the catalyst if the C₂+ is within the guaranteed limit. However, the SMR and WGS equilibrium temperatures can be calculated and compared with the observed temperatures.

Mass balance across the pre-reformer. Case 1—1,000 kmol/hr of treated (NG + recycled H₂)—has the following composition:

• H₂ = 2%
• N₂ = 5%
• CH₄ = 90%
• C₂H₆ = 2%
• C₃H₈ = 1%.

The inlet S/C ratio (mol/mol) is 3, the outlet pressure is 24.2 bara (23.9 atma), the inlet temperature is 450°C, and the outlet temperature is 400.2°C. The outlet composition (dry basis) reported by the laboratory is:

• H₂ = 18%
• N₂ = 4%
• CH₄ =73%
• CO = 0.03%
• CO₂ = 5%.

The unknowns include the outlet dry flow rate and the outlet steam, and it is not known how much steam has been consumed in the reactions to make products.

The inlet and outlet moles for Case 1 are shown in TABLE 1. In this example, d is the outlet dry flow rate, and s represents the moles of steam at the outlet. To calculate d using carbon balance, Eq. 10 is used:

900 + 20 × 2 + 10 × 3 = 0.73d + 0.05d + 0.0003d                                                                           (10)
   d = 1,243.11 kmol/hr

To calculate s using oxygen balance, Eq. 11 is used:

2,910 = 2 × 0.05 × 1,243.11 + 0.0003 × 1,243.11 + s                                                                      (11)
s = 2,785.3 kmol/hr

The total wet outlet flow (TABLE 2) is calculated in Eq. 12:

1,243.11 + 2,785.3 = 4,028.4 kmol/hr                                                                                                                                              (12)

To calculate the SMR equilibrium constant, Eq. 13 is used:

Kp (SMR)=P² × ([CO] [H₂ ]³)/[CH₄][H₂O] = 23.9² × (0.000093 × 0.056³)/(0.225 × 0.691) = 6 × 10^–5                                                 (13)

Eq. 13 can be substituted with Eq. 14 to obtain equilibrium temperature in °K²:

Ln (1/Kp) = 0.2513Z⁴ – 0.3665Z³ –
   0.58101Z² + 27.1337Z – 3.277                                                                                                                                                     (14)

where:

Z = (1,000/T)^–1—T is in °K
Ln (1/Kp) = Ln (1/6 × 10^–5) = 9.721

The equation can be solved using Excel: Z = 0.485; T_eq = 673.4°K (400.2°C). Therefore, the approach to SMR equilibrium in the pre-reformer is measured outlet temperature – equilibrium temperature = T − T_eq = 400.2 – 400.2 = 0°C.

A similar balance is done in Case 2 for treated naphtha feed. Case 2 feed is 80 kmol/hr of naphtha (estimated C number of 5.586), an S/C ration of 3 mol/mol, an outlet pressure of 24.1 bara (23.8 atma), an inlet T of 450°C and an outlet T of 477°C. The outlet composition (dry basis) reported by the laboratory was the following:

• H₂ = 21.5%
• CH₄ = 54%
• CO = 0.5%
• CO₂ = 24%.

The unknowns include the outlet dry flowrate and the outlet steam, and it is not known how much steam has been consumed in the reactions to make products.

To reiterate, d is the outlet dry flowrate, and s are the moles of steam at the outlet. The inlet and outlet moles for Case 2 are shown in TABLE 3. To calculate d using carbon balance, Eq. 15 is used:

446.88 = 0.24d + 0.005d + 0.54d                                                            (15)
d = 569.3 kmol/hr

To calculate s using oxygen balance, Eq. 16 is used:

1340.64 = s + 2 × 0.24 × 569.3 + 0.005 × 569.3                                       (16)
s = 1,064.53 kmol/hr

The total wet outlet flow (TABLE 4) is calculated using Eq. 17:

569.3 + 1064.53 = 1,633.83 kmol/hr                                                          (17)

To calculate the SMR equilibrium constant, Eq. 18 is used:

Kp (SMR) = P² × ([CO] [H₂]³)/[CH₄][H₂O] =
   23.8² × (0.00174 × 0.0749³)/(0.1882 × 0.6517) =
   3.38 × 10–3                                                                                          (18)

Eq. 18 can be substituted with Eq. 19 to obtain the equilibrium temperature in °K²:

Ln (1/Kp) = 0.2513Z⁴ – 0.3665Z³ – 0.58101Z² +
   27.1337Z – 3.2770                                                                                (19)

where:

Z = (1000/T)^–1—T is in °K
Ln (1/Kp) = Ln (1/3.38 x 10^–3) = 5.691

The equation can be solved using Excel: Z = 0.333; T_eq = 750°K (476.9°C). Therefore, the approach to SMR equilibrium in the pre-reformer is T − T_eq = 477 − 476.9 = 0.1°C.

Part 2

Part 2 will be featured in the November issue of Hydrocarbon Processing. HP

LITERATURE CITED

1. Riazi, M. R., Characterization and Properties of Petroleum Fractions, ASTM International, West Conshohocken, Pennsylvania, 2005.
2. Twigg, M. V., Catalyst Handbook, Second Edition, CRC Press, Boca Raton, Florida, 1996

The Author

Related Articles

From the Archive

Comments