H2S is toxic at 10 ppm, entirely lethal at 800 ppm, highly corrosive to equipment, flammable when in excess of 4.3% by volume in air, and unpleasantly odorous at a threshold of less than 1 ppb.
Unfortunately, H2S occurs abundantly in the world’s fossil fuel reserves. The sulfur recovery unit (SRU) of a refinery is dedicated to processing the H2S stripped from the hydrocarbon fuel through a series of operations that convert it into water and harmless elemental sulfur, which can be sold and repurposed in fertilizer, gunpowder, and more.
The Claus process is the industry standard for treating the H2S-rich “sour” gas. In a furnace, H2S is combusted:
3H2S + 3⁄2O2 SO2 + H2O + 2H2S
A catalytic converter reacts the products of the combustion to create elemental sulfur in various crystalline forms:
2H2S + SO2 2H2O + 3⁄XSX
As can be deduced from the second reaction above, the typical Claus reaction runs most efficiently when the stoichiometric ratio of H2S to SO2 is controlled at 2:1. The 1st reaction above demonstrates that this ratio is controlled by adjusting the amount of available oxygen.
The analyzer uses a high resolution UV-Vis spectrophtometer to acquire a 200-800 nm absorbance spectrum of the sample gas. Within this spectrum, the analyzer can identify the distinct absorbance curves of each analyte, measure the height of these curves, and correlate that value directly to real-time concentration.
The optical assembly of the TLG-837 is depicted below, illustrating the complete path of the signal.
The signal originates in the pulsed xenon light source and travels via fiber optic cable to the flow cell disk, which is built into the head of the probe. Passing through the length of the flow cell, the signal picks up the absorbance imprint of the continuously drawn sample gas.
Exiting the flow cell on the opposite end, the signal travels by fiber optic cable to the spectrophotometer, where a holographic grating separates the signal into its constituent wavelengths, focusing each wavelength onto a corresponding photodiode on a 1024-diode array. This is known as dispersive spectrophotometry.
A conventional ‘multi-wave’ photometer measures a chemical’s absorbance at one pre-selected wavelength with one photodiode. This ‘non-dispersive’ technique uses an optical filter or line source lamp to remove all wavelengths but the pre-selected measurement wavelength.
By contrast, the TLG-837 uses a dispersive spectrophotometer to acquire a full, high-resolution spectrum. Each integer wavelength in the spectral range is individually measured by a dedicated photodiode.
A single photodiode is susceptible to noise and signal clipping. As accepted in the lab community for decades, the only way to eradicate this source of error is to use many photodiodes measuring at many wavelengths. Compiling the data from all these photodiodes produces an absorbance spectrum instead of a single data point:
The internal components of the TLG-837 analyzer unit are indicated below (door removed):
The HMI controlling the spectrophotometer and communication provides a simple, touch-screen visual interface. Running our proprietary ECLIPSE software, the HMI offers the user several display choices (e.g. standard numeric display, trendgraph, bar graph).
From this interface, the user can quickly perform tasks and adjust settings, including:
The heart of the TLG-837 is the diode array spectrophotometer. This device contains the light source as well as the detector which measures the absorbance spectrum.
A highly evolved device, the nova II has several distinctive features which allow it to excel in tail gas analysis:
Our fibers are all manufactured in-house to ensure spectroscopic-grade quality. The stainless steel cladding provides proven durability in the field. Before shipment, each fiber is tested to ensure it meets transmission benchmarks, Exceptional UV light transmission is achieved through our presolarization technique.
The fibers connect to the flow cell through rugged steel collimators, and are thus not wetted to the sample fluid. Optional cooling extensions provide further protection from hot samples.
This device receives digital information from the HMI and converts it to 4-20 mA analog signals, the industrial standard for communication with the DCS.
The HMI runs ECLIPSE software off of a solid state drive (SSD) with ample storage for saving historical measurement data. Spare SSDs have the ECLIPSE software preloaded and can easily be swapped in with minimal downtime.
The ECLIPSE software is able to measure H2S, SO2, COS, and CS2 simultaneously by de-convoluting the absorbance curve of each analyte from the total sample absorbance structure. This method is unique, as it does not require physical wavelength isolation. While other systems utilize moving parts (e.g. filter wheels) or multiple line source lamps (each requiring replacement), the TLG-837 uses the power of rich data and adheres to a simple, solid state design with a single, long-life light source.
Multi-component spectroscopy is made possible by the principle of additivity: according to Beer’s law, the absorbance of a mixture at any wavelength is equal to the sum of the absorbance of each chemical in the mixture at that wavelength.
The TLG-837 measures the total absorbance curve of the tail gas and solves for the individual absorbance curves of the analytes by using a matrix of equations. As illustrated above, each photodiode-wavelength combination supplies a single equation to the matrix, in the form:
A’(x+y) = A’x + A’y = e’xbcx + e’ybcy
Where A’ is the absorbance at wavelength ‘, e’ is the molar absorptivity coefficient at wavelength ‘, c is concentration, and b is the path length of the flow cell.
Under upset conditions, the H2S/SO2 ratio may deviate widely from the desired 2:1 optimization point; also, some modified Claus processes operate with a controlled ratio much higher than 2:1. A high-performing tail gas analyzer must therefore sustain accuracy when the ratio is outside the expected range (‘off-ratio’ conditions).
In a multi-wave photometer, the response to concentration change is limited by a single photodiode’s ability to measure swings in absorbance. That lone diode is highly susceptible to noise and signal clipping at out-of-range absorbance. The TLG-837 spectrophotometer overcomes this constraint by using a 1,024-diode array; the full spectrum has regions of measurement wavelengths that are suited to different scenarios, while statistical averaging of each diode’s reading serves to eradicate noise.
By virtue of full spectrum acquisition, the TLG-837 sustains specified accuracies when H2S/SO2 ratio reaches as high as 100:1 or as low as 1:20. This dynamic range is unrivaled in tail gas analysis.
Applied Analytics’ patented TLG-837 DEMISTER Probe is mounted directly on the process pipe. The physical measurement occurs inside the probe head, minimizing the sample transport time. Entrained sulfur vapor is removed from the sample as an internalized function.
This probe was designed to be lightweight and compact, so it’s easy to install and service by a single technician.
Tail gas contains elemental sulfur which is quick to condense and plug mechanical cavities or obstruct optical signals. The DEMISTER Probe removes sulfur from the rising sample as an internalized function within the probe body. Recycling the steam generated by the Claus process, the probe controls the temperature along its body at a level where all sulfur vapor in the rising sample condenses and drips back down to the process pipe.
Inside the probe, an internal ‘demister’ chamber (concentric to the probe body) is fed with low pressure steam (see E & F). Since the LP steam is much cooler than the tail gas, this chamber has a cooling effect on the rising sample.
Elemental sulfur has the lowest condensation point of all of the components in the tail gas. Due to the internal probe temperature maintained by the LP steam, all of the elemental sulfur in the rising sample is selectively removed by condensation while a high-integrity sample continues upward for analysis in the probe head.
The point of interaction between the light signal and the sample gas occurs in the flow cell disk inside the probe head (C & D). The flow cell disk has a built-in HP steam channel (A & B) to heat the cell and ensure that any present sulfur remains gaseous—eliminating the possibility of condensation on the optical windows.
An aspirator (G) creates a Venturi effect which pulls the sample up the probe body intake path, through the flow cell for analysis, and down the return line. The used sample is released back into the process pipe (H).
Applied Analytics design centers on inherent safety. The major safety flaw of other tail gas analyzers is that they bring the toxic sample fluid into the analyzer enclosure for analysis. Not only does this practice expose the system electronics to higher corrosion effects, it also poses a lethal threat: if there is any leak in the instrument — especially inside a shelter — the human operator is placed at enormous risk.
The key difference between the TLG-837 and other tail gas analyzers is the use of fiber optic cables: we bring the light to the sample instead of bringing the sample to the light. The toxic sample only needs to circulate through the probe, and never enters the analyzer electronics enclosure.
In order to regulate the pressure of the steam going to the DEMISTER Probe, the user can build their own panel or purchase the optional TLG-837 Utility Control Panel (UCP). Standard functions of the UCP include:
With the features above, the UCP is a standardized panel engineered for turnkey integration. Note: no part of UCP is wetted to sample.
The TLG-837 only requires a one-time calibration during installation. Designed for unattended operation, the system depends on Auto Zero to maintain accuracy. This automated task normalizes the spectrophotometer reading while running a zero-absorbance gas (e.g. nitrogen) through the flow cell.
When Auto Zero initiates (following a user-defined schedule), the ECLIPSE software automatically operates the appropriate valves via relays to purge the flow cell with zero gas and save a new zero spectrum.
In a typical usage profile, Auto Zero is set to run every 8 hours. The task requires approximately 120 seconds during which the measurement output is frozen. Under these settings, the TLG-837 can provide greater than 99.5% analyzer uptime.
|Measurement Principle||Dispersive ultraviolet-visible (UV-Vis) absorbance spectrophotometry|
|Detector||nova II™ UV-Vis diode array spectrophotometer|
|Spectral Range||200-800 nm|
|Light Source||Pulsed xenon lamp (average 5 year lifespan)|
|Signal Transmission||600 μm core 1.8 meter fiber optic cables
Other lengths available
|Sample Introduction||In situ DEMISTER Probe|
|Analyzer Calibration||Calibrated with certified calibration fluids; no re-calibration required after initial calibration; measurement normalized by Auto Zero|
|Verification||Simple verification with samples|
|Human Machine Interface||Industrial controller with touch-screen LCD display running ECLIPSE™ Software|
|Data Storage||Solid State Drive|
|Analyzer Environment||Indoor/Outdoor (no shelter required)|
|Ambient Temperature||Standard: 0 to 35 °C (32 to 95 °F)
With optional temperature control: -20 to 55 °C (-4 to 131 °F)
To avoid radiational heating, use of a sunshade is recommended for systems installed in direct sunlight.
|Standard Outputs||1x galvanically isolated 4-20mA analog output per measured analyte
5x digital relay outputs for indication and control
1x K type ungrounded thermocouple input
|Optional Outputs||Modbus TCP/IP; RS-232; RS-485; Fieldbus; HART|
|Wetted Materials||Stainless Steel 316/316L, Kalrez
Other materials available
|Analyzer Enclosure||wall-mounted NEMA 4X stainless steel type 304 Enclosure
Other enclosures available
|Probe Material||Stainless Steel 316/316L
Other materials available
|System Dimensions||Analyzer: 24” H x 20” W x 8” D (610mm H x 508mm W x 203mm D)
Probe (Average): 36” length x 12“ widest diameter (914mm x 305mm)
|System Weight||Analyzer: 32 lbs. (15 kg)
Probe (Average): 29 lbs. (13 kg)
|H2S||0-2%||± 1% full scale||± 0.4%|
|SO2||0-2%||± 1% full scale||± 0.4%|
|air demand||user-defined||± 1% full scale||± 0.4%|
|Off-Ratio Range||100:1 H2S:SO2 20:1|
|Response Time||1-5 seconds|
|Zero Drift||±0.1% after 1hr warm-up, measured over 24hrs at constant ambient temperature|
|Sensitivity||±0.1% full scale|
|Noise||±0.004 AU at 220 nm|
|Standard Design||General Purpose|
|Available Options||ATEX, IECEx, EAC|
|Please inquire with your sales representative for additional certifications (CSA, FM etc.).|