Separation of Three Phases, Gas, Liquid and Solid Using a Cyclone Injected with Hot Hydrogen.

[NNadir]

The paper I'll discuss in this post is this one: Hydrocyclone Settler (HCS) with Internal Hydrogen Injection: Measure of Internal Circulation and Separation Efficiencies of a Three-Phase Flow (Roberto Galiasso Tailleur, Andres G. Peretti, Ind. Eng. Chem. Res. 2020, 59, 3, 1261-1276) Dr. Tailleur is a Chemical Engineer whose affiliation in the paper is listed as Simon Bolivar University in Miranda Venezuela. Venezuela, for those who do not know, is a large producer of the dangerous fossil fuel petroleum, which is toxic in not only environmental and epidemiological sense, but is also toxic in an economic and political sense. This has certainly been true in Venezuela. Venezuela is a case in point that political and economic absolutism on the far left is not particularly less odious and less destructive than political and economic absolutism on the right. The main product of the country these days is economic refugees.

This paper is about the processing of dangerous fossil fuels, and concerns the catalytic alkylation of C2 and C3 alkenes, ethylene and propene, with isobutane, presumably to make the dangerous fuel gasoline.

As an opponent of the use dangerous fossil fuels, on the grounds that they are destroying a future that does not belong to us, that it is not our right to destroy, it may seem strange that I am as interested as I am in this kind of technology. Nevertheless I have been recently emphasizing that many of the technologies in industries that should be abandoned on the grounds they are not sustainable, have use in other industries, or in new uses for expansion of applications of extant industries. Indeed this was historically true of the dangerous petroleum industry. The original purpose for the distillation of dangerous petroleum which was industrially pioneered by John D. Rockefeller, was to make lamp oil to replace whale oil as the over hunting of whales was leading to the decline of the species that had nothing to do intrinsically with pollution, making whale oil expensive and difficult to obtain. The development of the distillation process led to a search for what was then a by product of distillation, gasoline. The destruction of the planetary atmosphere followed in the following century and a half.

I have been making this point as well about the useless solar thermal industry, another so called "renewable energy" industry dependent upon turning huge pristine ecosystems into industrial parks that have low capacity utilization, in this case desert ecosystems. The solar thermal industry will never be able to make economically viable hydrogen to replace the source of more than 98% of the current source of the world's hydrogen, dangerous natural gas and dangerous coal, but the technologies explored in papers around thermochemical water and carbon dioxide splitting cycles that are often described in papers as being applicable to solar thermal technology are equally viable to sustainable and low environmental impact nuclear systems.

According to the reference in the paper, the compounds in the three phases described in this paper for which this hydrocyclone was developed are as follows: The solid is a platinum sulfate-titanium-zirconium catalyst supported on silica, SiO2, basically sand. The gases are isobutane, propene and butene, as described above, and the liquid is the condensate, dangerous alkanes that are components of dangerous gasoline.

Be this as it may, I am always thinking about ways that future generations will clean up the mess that our generation produced in a kind of drunken sybaritic ecstasy of material consumption, albeit the ecstasy in question being distributed among parties in an ever less judicious way.

The biggest mess we leave of course, is a destroyed atmosphere. Among the engineering routes to removing the dangerous fossil fuel waste carbon dioxide from the atmosphere, all of which will be challenging, I personally regard reformation of biomass with supercritical water, supercritical seawater, or supercritical carbon dioxide to be potentially viable routes. These technologies will involve three phase (actually four phase) separations, and thus my interest in the technology. Many, most, if not all, will involve hot hydrogen.

From the introduction of the paper:

Figure 1 shows the new alkylation–regeneration process scheme that is described in detail in refs (1) and (2); it consists of a slurry transport reactor (STR, used for alkylation), two stages of gas–solid–liquid separation in a hydrocyclone stripper (HCS1 and HCS2), and a fluidized bed reactor (FBR, employed for catalyst regeneration). The HCS of Figure 1 is used to separate gases, liquids, and solids from the stream that left the alkylation reactor at 360 K and 1.4 MPa. It delivers solids to the FBR and gases and liquids to other separation stages.(1) The solid obtained (spent alkylation catalyst) in the two HCS stages must be stripped of adsorbed hydrocarbons by hot hydrogen before being fed to a fluidized bed catalyst regenerator. Ninety-five percent of the regenerated catalyst is sent back to the STR. Catalyst purge and make up is a function of catalyst deactivation, HCS efficiency, and fines produced in the process.

Figure 1:



The caption:

Figure 1. Hydrocyclone–settler system.

The introduction continues:

The feed of the HCS is composed of hydrogen soluble in light alkylate and 8% by weight of the catalyst; the latter contains less than 2% of 1–10 ?m and more than 97% of 40–60 ?m in particle diameter. More than 98% of the solids and less than 10% of liquids must be recovered at the underflow (UF) stream. In this process scheme, two stages are used to obtain the maximum recovery of solids and liquids and deliver preheated solids into the regenerator (Figure 1).
Hydrocyclones have been successfully used in the industrial separation of solids for more than 40 years. The design is simple, easy to operate, and of low operating and maintenance costs; these devices are very important to perform solid separations; nevertheless, high solid efficiency in hydrocyclones is difficult to achieve when there are small differences between liquid and solid densities in the feed and they contain very fine particles. In conventional (isothermal) liquid–solid separation, the average cut size dp50 is related to the inlet pressure at an order of 0.25 and to the hydrocyclone diameter, according to Bradley(2) and Rietema(3) equations...

...Previous experiments with a small hydrocyclone and the study done for the spouted bed reactor development were used to select current base case (BC) dimensions and operating conditions for the new hydrocyclone settler (HCS1). Then, the HCS1 was tested to mainly explore the effect of lower cone lengths, feed and hydrogen flow rates, and temperatures using five types of sensors: ECT, differential pressure, pressure, and temperature as well as by sampling. In addition, other amounts and particle size distributions were used to compare with BC. These sensors were calibrated using a well-known flow vessel at similar operating conditions to those of the HCS; data were consolidated, and the methodology was used in the current study of HCS1.
There are several characteristics of this device that are not mentioned in the literature about conventional three-phase cyclones:

(1) radial and axial gas, liquid, and solid heating by hot hydrogen,

(2) vaporization of hydrocarbons,

(3) slurry lift in the riser by effects of hydrogen injection through a nozzle,

(4) tangential and axial flows of gases in the riser that have a divergent outlet,

(5) different types of gas cores (formed hydrogen plus vaporized hydrocarbons),

(6) gas core positions controlled by the settler level of wet solids,

(7) tailpipe discharge into a settler with controlled levels of solids, and

(8) use of high-pressure metallic hydrocyclones with the flow, level, and pressure controlled by three automatic loops.

The main objective of this device is to maximize the separation of coarse (>10 microns) and minimize that of fines particles (<10 microns) and liquids going into the regenerator operating at a 1.4 MPa inlet pressure; a secondary objective is to preheat the catalyst for regeneration in the fluidized bed, and the third one is to minimize the delta of pressure, erosion, vibration, and pressure oscillation in a steady-state operation.

The streams were characterized by their particle size distribution, pressure, temperature, amount and type of solids, and gas and liquid content. Gas–liquid equilibria were calculated using the Peng–Robinson state equation.

More than 1800 experimental points were used to calibrate the sensors. The results are reported in the Supporting Information. A total of 2800 points were obtained with HCS1 using nine high-pressure prototypes. The separation efficiencies and internal circulation were measured, and the results were compared to the values predicted by known published equations.

(Yesterday I attended a wonderful lecture, in the context of the development of understanding fusion plasmas, on the use of "artificial intelligence" in the processing and weighting of fairly extreme multivariate analytical inputs, which I might imagine would have application for a system of measurements for this dangerous fossil fuel technology. It was unbelievably fascinating. It may be available as a video at the above link in a few weeks.)

Figure 2 in the paper shows the types of analytical tools used in the analysis of this device:



The caption:

Figure 2. (a) Global scheme of slurry preparation and HCS1 and HSC2 with secondary hydrogen preheating and injection at the inlet of riser. (b) Five ECT sensors located in the cone with capacitance, temperature, and pressure detectors and an imaging processor system; (c) hydrocylone dimensions; and (d) list of sensors (see location in (a)) connected to a data logger and processing computer.

Figure 3 shows the types of readouts being processed:




The caption:

Figure 3. Different variables plotted as functions of operational time (minutes). Upper part: inlet feed and hydrogen mass flow rates. Middle part: delta of pressure and gas temperatures at the outlet. Lower part: slurry mas flow rates at the UF and OVF streams. HCS1 at base-case operating conditions.

(These inputs are considerably of lower dimensionality than the measurements of fusion plasma devices, but depending on the time resolution can still be quite complex.)

Two tables describing the inputs and dimensions of the pilot device:





Some overview remarks of the conduct of experiments:

The pilot plant was continuously operated, and the flow rate, temperature, vibration, and pressure were recorded (Figure 3); mass balances were performed every 10 min. The results are reported in Table3 as an example. Internal circulation obtained by the sensors is shown in Figure 4, and the radial profiles of pressure, temperature, and solid content are in Figure 5. The flow in the apex and tail pipe is depicted in Figures 6 and 7, and the effect of apex and vortex finder diameters is in Figure 8.

Table 3:



The figures:




The caption:

Figure 4. (a) Half-HCS1 with the spires going downward (yellow) by the wall and going upward (green) around the gas core. (b) Radial distribution of solids (ECT) at z = 0.5 and z = 0. (c) Profiles of temperature (red), vaporization (purple), solid concentrations (green), and pressure near the external wall. (d) PSD in the feed and OVF and UF streams (%, with respect to the stream).

The caption:

Figure 5. (a) Radial profile of temperature (red, thermocouple) and the core radius (yellow, ECT); (b) radial distribution of pressure; (c) radial distribution of delta pressure; and (d) ECT measure of solid distribution. Operating conditions for HCS1: FH2 = 0.4 kg/s, TH2 = 560 K, Lc1 = 0.5 m, uo1 = 4 m/s, Lc1 = 0.4 m, and Do/Dc = 0.4.

The caption:

Figure 6. (a, b) ECT spaced by 8 cm in the connecting (tail) pipe during circulation (circ) and spray modes of discharge. (c) Twin-plane Cs/Cs,av ratio measured by ECTV as a function of delay time for circulating and circulating-spray-circulating types of discharge in the tail pipe.

The caption:

Figure 7. Effect of hydrogen in the oscillation of the core and frequency of circ and circ-spray-circ mode of solid discharge. (a) ECT at z = 0.5 and z = 0 as a function of time; (b) amplitude of the vibration at the apex (10 kHz) as a function of time.

The caption:

Figure 8. Effect of Du and Dr in coarse solid separation efficiency at optimal hydrogen flow rates. Delta of pressure of vortex–apex of ?4 ± 0.5 kPa and vibration frequency of 20 ± 4 Hz.

The caption:

Figure 9. (a) Effect of PSD in the feed in solid distributions at the outlet streams. Feed A (black dashed lines), UF (red dashed lines), and OVF (green dashed lines). Feed B (violet dashed lines), UF (gray dashed lines), and OVF (blue dashed lines). (b) 2D ECD image of radial particle distribution at z = 0.5 for A and B solid distributions in the feed. (HCS1 base-case dimensions and operating conditions.)

Some discussion of the results:

The main difference between HCS1 and previous hydrocyclone technology is the use of hot hydrogen that “complements” the effects of centrifugal force in the separations of gas, liquid, and solid from the stream that leaves the alkylation reactor. Both liquid and solid separations (efficiencies) need to be maximized because they affect the economy of the process.
Sensors (ECT, differential of pressure, pressure, and temperature) and sampling allow determining the flow pattern of solids in different areas of the hydrocyclone. The tests found that hydrogen produces a different type of fluid circulation in the lower cone, riser, and apex than that reported for conventional hydrocyclone, desander, and deoiling devices. Hydrogen produced important hydrocarbon vaporization that changes slurry properties (density and viscosity), radial and axial distributions of solids, temperature, and delta of pressure and produces a carry-over of slurry through the riser. Hydrogen injection is responsible for pressure oscillation and additional vibration of the HCS.

The simulation of HCS1 using published correlations show important deviation. For example, the calculation presents a deviation higher than 38% in the prediction of for HCS1 operating with cold hydrogen (363 K and 1.4 MPa), but the differences are higher when using hot (530 K) hydrogen. Reynolds numbers of the feed (inlet), required to produce high solid/liquid separation at the apex in HCS, are half those measured by other authors in isothermal hydrocyclones (see, for example, the value used by Wang and Yu(30)); these authors observed a shorter reverse core in flooded hydrocyclones at almost twice the inlet Re number needed for solid separation than that calculated for the isothermal HCS1 device. The HCS operates at a relatively low inlet Re number with a shorter cone and broader vortex finder and underflow diameters than those of conventional isothermal hydrocyclones. There is no correlation between d50 and the inlet Re number as it was for conventional devices (see Gu and Liow(39)).

An excerpt of the authors conclusions:

________________________________________
A total of 2800 experimental points have been obtained for HCS1 to determine the best dimensions and operating conditions for gas, solid, and liquid separation. Stability, efficiency, delta of pressure, catalyst attrition, and the effect of hydrogen flow and temperature were studied using BC dimensions, selected based on previous studies; the effect of some critical dimensions were studied by departing from BC. The objectives of the separations are imposed by the economy of the process. Minimum vibration and pressure fluctuations and pressure losses are operational requirements. The results demonstrate that

(1) the solid separation efficiency increases sharply from fines (0–10 microns) to coarse particles (40–70 microns) as expected. The efficiency for fine-particle separation is higher than that observed in conventional hydrocyclones.

(2) The solid separation mainly occurred in the shorter than conventional lower cone where coarse particle tangential and radial velocities are accelerated by centrifugal forces and hydrocarbon vaporization. There is a nonlineal radial profile of temperature, pressure, and solid concentration across and along the lower cone. Without hot hydrogen injection, there is not enough centrifugal forces to separate solid and liquid for BC dimensions.

(3) In steady-state conditions, the slurry, high-in-solid, moves downward, rotating against the internal wall in the lower cone, axis, and tail pipe. The level of solids at the settler ropes the discharge and induces the upward movement of low-in-solid slurry that helps the separation at the apex and seals the bottom of the gas core...

Additional points are made in the conclusion, and the paper features extensive discussion of these engineering parameters.

These sorts of things, I know, are very esoteric, but these are the types of things about which future generations will need to know to address the consequences our irresponsibility.

Enjoy your Sunday.