This application is not a continuation in part. It is an improvement over, and related to, my earlier filed applications listed below, which are incorporated by reference:

DCKT USSN TITLE

08/217821 filed on March 25, 1994.

7579 08/367500 TWO PHASE TREATMENT OF GAS TO REMOVE HALOGENS

7580 08/367498 REMOVAL OF ACIDIC HALIDES FROM GAS

STREAMS

7581 08/367501 TWO PHASE REMOVAL OF HALIDES FROM LIQUID

HYDROCARBONS

7582 08/367411 THREE PHASE REMOVAL OF HALIDES FROM

LIQUID HYDROCARBONS

7583 08/367499 REMOVAL OF ACIDIC HALIDES FROM HOT GAS

STREAMS AND ATTRITION REGENERATION OF CAUSTIC

7584 08/367412 DISSOLVING SALT ON SOLID CAUSTIC WITH

OIL

7585 08/367413 CHEMICALLY ACTIVE VAPOR/LIQUID SEPARATOR

Field of the Invention

This invention relates to removal of halogens, especially chlorides, from liquid and vapors such as refinery streams.

Brief Summary of the Invention Accordingly, the present invention provides a process for removing acidic halides from a gas stream or liquid

hydrocarbon stream comprising contacting a dry stream containing acidic hal-ogen compounds with particles of an inert porous support having a solid caustic phase in a portion of the particle, neutralizing at least a portion of said acidic halogens by reaction with said solid caustic phase to form salts which deposit on the surface of said solid caustic and within said porous support, removing from contact with said particles a treated hydrocarbon stream having a reduced content of acidic halogens. In another embodiment the invention provides a method of regenerating an adsorbent used to remove acidic components from vapor or liquid streams, said adsorbent comprising a caustic impregnated and salt contaminated inert carbon based support, comprising washing salt from said adsorbent by contact with water present in an amount sufficient to dissolve at least a portion of said salt contamination to produce washed adsorbent with a reduced salt content and a residual caustic content, drying said washed adsorbent to produce adsorbent with a residual caustic content which is used to remove said acidic components.

In another embodiment the invention provides a method of rejuvenating an adsorbent used to remove acidic components from vapor or liquid streams, said adsorbent comprising a caustic impregnated and salt contaminated inert carbon based support, comprising washing salt from said adsorbent by contact with water present in an amount sufficient to dissolve at least a portion of said salt contamination to produce washed adsorbent with a reduced salt content, drying said washed adsorbent to produce dried adsorbent; and impregnating said dried adsorbent with an aqueous caustic solution to produce rejuvenated adsorbent which is used to remove said acidic components.

In another embodiment the invention provides an adsorbent for removal of acidic components from a flowing stream comprising particles of activated carbon or charcoal having an internal pore structure and containing at least 20

weights of caustic per 100 weights of carbon or charcoal, and wherein said caustic clogs at least a portion of said pores with a caustic continuous phase.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows chloride removal v. chloride loading for four different adsorbents.

Figure 2 shows chloride removal v. chloride loading through five regeneration cycles.

Detailed Description of Preferred Embodiments

The process and composition of the present invention may be used wherever conventional alumina treaters or the like have been used to remove acidic components from gas or liquid streams.

Such treaters have been used to treat dry liquid streams such as reformate, and dry gaseous streams such as reformer recycle gas and off gas from the reformer or gas streams resulting from regeneration of the catalyst. All streams treated should be relatively dry, i.e., free of any separate water phase or have a low water vapor partial pressure. This is because the presence of liquid water, or large amounts of water vapor in a gaseous stream, could dissolve the caustic. The problem of too much water is no different than that encountered when using solid alumina treaters.

SOLID INERT SUPPORT

The process and composition requires use of a solid inert support to hold the "solid" caustic.

Charcoal, activated carbon, coal, silica, and the like may all be used. The support should be essentially free of any catalytic activity, both alone and after impregnation with caustic and after use in removing acidic halogens from the process stream.

Alumina is not a good support. Alumina can react with caustic, forming sodium aluminate or other reaction products after caustic is added. Alumina is also unsuitable in that it can react with, e.g. , chlorides in the flowing stream to form aluminum chloride which can form organic polymers from hydrocarbons and organic chlorides.

Phrased another way, the support has to do very little except be there as a convenient inert place which may be repeatedly subjected to alkaline impregnating solutions. It is important that the support not react substantially with the impregnated caustic, which might affect the integrity of the support and result in some caustic consumption in advance of acidic halogen removal.

The most cost-efficient use of the support is to add enough caustic to the support to create solid caustic inside a porous support, hence my term clogged.

If the charcoal is the heart of a chloride removal process, my process works best when most of the heart's passages are clogged. Unless at least some of the passages are clogged, it is not possible to take advantage of the salt building capacity I observed with solid beads of caustic, and the bed will not have sufficient endurance to make for a viable process.

The most efficient supports in terms of caustic loading are also efficient chloride removers. The preferred material is a vapor phase activated carbon such as Atochem. Atochem vapor phase activated carbon, can be impregnated one time with 50 wt % NaOH solution to load 20 to 25 weights of NaOH per 100 weights carbon, hereafter WPHC or Weights per Hundred Weights Carbon. Other types of carbon do not take up as much caustic during such an impregnation procedure, as an example, tests were run with Centaur carbon, and this material showed an NaOH uptake of about 10 WPHC. Preferred materials for use herein will have a caustic uptake during impregnation with a 50 wt % NaOH solution in

excess of 10 WPHC, more preferably in excess of 15 WPHC, and most preferably in excess of 20 WPHC.

ALKALINE COMPONENT Most refiners will use NaOH because of ready availability and low cost, but other forms of water soluble alkaline materials such as glassmakers alkali, KOH, CaO, MgO and the like may be used, either alone or in mixtures.

The invention has little to do with the type of alkaline material used - refiners have been using these materials for over a century - but rather with a unique way of placing this material on a porous support.

SUPPORT IMPREGNATION It is beneficial to use multiple impregnations of the solid support with a suitable caustic solution, or some sort of impregnation procedure which adds enough of the desired alkaline component to form a continuous phase of solid alkaline material in at least a core portion of the support. It is believed that porous supports, such as alumina, were routinely impregnated with caustic solution just until such a solid phase formed. Many alumina treaters contain perhaps 5 wt % NaOH.

In my composition, the impregnation is preferably repeated until much of the interior structure is clogged with solid caustic. The carbon or other inert support should contain from 5 to 200 wt % NaOH (or 5 to 200 PHC) . In experiments with charcoal it was easy to add at least 20 wt % NaOH per impregnation, with repeated impregnations it was possible to add have carbon containing 20, 40, 60, 80 and 100 wt %. Each impregnation added roughly 20 to 25 wt % NaOH.

Even with high caustic loadings essentially all NaOH is in a form where it could be used up to react with chlorides in the flowing stream to be treated.

TREATING PROCESS

My process is simple. The stream to be treated for acidic halogen removal contacts solid caustic in a porous inert support. No aqueous phase is present nor are any chemicals added except for the initial load of solid caustic in the inert support.

Even the chemistry of my process is simple. Simple neutralization reactions are involved which proceed rapidly. The primary reactions involved when, e.g., chlorides are removed are:

HC1 + NaOH - NaCl + H ₂ 0 NH ₄ C1 + NaOH - NH ₃ + NaCl + H _? 0 FeCl ₃ + NaOH - Fe(OH) ₃ + ₃ NaCl + ₃ H,0 The reaction products are water and salt. The water is present in such small amounts that it may remain dissolved in the process stream being treated. The salt deposits build up in the charcoal or other porous support.

The process may be continued a long time, until acidic component removal drops off. Many refiners will simply dump their carbon beds and replace with fresh material, but many will want to take advantage of regeneration and rejuvenation procedures which were developed. Regeneration simply involves pumping water through the bed to remove salt deposits. Rejuvenation involves washing followed by re- impregnation with fresh caustic. Both procedures are discussed in more detail below.

REGENERATION - SALT REMOVAL

Most refiners and chemical plant operators will prefer to use a support with caustic loadings of 25 - 50 - 100 or more PHC. Such caustic clogged supports will lose effectiveness over time as salt deposits build up within the support on top of the solid caustic phase. Such salt deposits can be removed, and the charcoal regenerated, by washing the salt clogged support with controlled amounts of water.

The salt comes off rapidly. Usually 3 or 4 washings with water - with a washing defined as pumping 5 bed volumes of water over the bed - remove sufficient salt to restore the activity of the carbon bed. Surprisingly, the wash water is only slightly alkaline after the first washing, and becomes even less alkaline in subsequent washings. Thus the pH of the initial effluent never exceeded 9, and the pH of effluent of subsequent washings was even less. Thus the utilization of NaOH loaded on the carbon for HCl reaction was essentially complete.

Surprisingly, such repeated washing with water never seemed to remove caustic. As the test results, presented in the EXPERIMENTS section show, both the pH of the wash water and the NaCl concentration decreased through 4 regeneration procedures. This selectivity for salt even in the presence of large amounts of NaOH helped avoid problems of exother icity and generation of steam which would have occurred had solid caustic rapidly dissolved in the wash water. The dissolution of NaOH is highly exothermic, while NaCl dissolution is slightly endothermic, so carbon loading of NaOH is an inherently safer, if not inherently safe, way to package caustic for refinery and petrochemical use.

After salt is removed the remaining solid caustic in the porous support may be used to remove additional amounts of acidic halogens from the flowing stream.

REJUVENATION PROCEDURE

The activity and capacity of the carbon bed may be significantly restored by the following rejuvenation procedure. the more caustic may be added by repeating the impregnation procedure.

The first step is removal or deposited salt, using the regeneration procedure. Then the carbon is preferably dried, and re-impregnated one or more times with concentrated NaOh solution. The process is never complete

so that the quantity of NaOH impregnated decreased gradually after each impregnation.

The activated carbon support provides bed structure which is stable during operation and during regeneration/rejuvenation. The carbon also provides the surface to deposit the NaOH so that it can be consumed nearly completely in the adsorption cycle.

EXAMPLES

The following examples represent actual laboratory experiments.

The synthetic gas used for testing in this study was obtained by passing N2 gas over a HCl reservoir to carry in HCl. This synthesis gas was fed downflow into the reactor above the solid caustic bed. The HCl concentration would drop during the long term tests, so the HCl concentration in the gas continued to decrease almost logarithmically. As the HCl content became too low, additional concentrated HCl was added to the HCl reservoir to keep the HCl content in the synthetic gas at the desired levels. To minimize the change in HCl content of the gas, concentrated HCl was used in the reservoir and covered with naphtha to control the diffusion rate and, in turn, the HCl content in the gas.

Four adsorbents were tested.

The first material was alumina impregnated with caustic. This was a sample of a commercially available adsorbent used at one of our refineries to remove acidic components from gas. This is a prior art material.

The second adsorbent was beads of NaOH. These were chemical grade semispheres with a diameter of about 0.5 cm. While not, strictly speaking, prior art, such materials are the subject of my prior patent applications.

The third material (Invention) was an Atochem vapor phase activated carbon. This was dried in a vacuum oven at 120°C. The dried carbon was impregnated with 50 wt % NaOH solution to the desired NaOH levels of 40, 60 and 80 %. The NaOH loading of 20 to 25 % can be achieved in one impregnation.

To increase the NaOH loading, the impregnation step has to be repeated.

The fourth material (invention) was Centaur carbon. The NaOH loading capacity of Centaur is smaller than the Atochem vapor phase carbon. The uptake of NaOH from 50 % NaOH solution per impregnation were 20-25 % and 10 % for Atochem vapor phase carbon and Centaur carbon, respectively.

The reactor system used a packed column. The diameter of the column was 14 mm and the bed height was 40.6 cm. The HCl containing gas passed downflow through the bed at a gas flow rate of 1.6 1/min and a gas velocity of 0.57 ft/sec. The adsorbents were tested side-by-side as shown in the following brief table.

The increase in HCl content was to speed up the test to make use of the higher capacity of the Atochem and Centaur carbon based adsorbents.

The synthesis gas and the effluent gas were scrubbed with NaOH solution and the chloride contents were determined using an Orion chloride electrode (Model 94-17B) . From this the HCl contents of synthesis gas and the effluent gas were calculated. These data were used to generate chloride removals shown in Figure 1. The chloride adsorbed on the bed was the difference in chloride contents between the synthesis gas and the effluent gas. The chloride uptake on adsorption by the bed was cumulated and calculated in terms of wt % of the adsorbent, namely:

Chloride adsorbed, wt % = (g of chloride/g of adsorbent) * 100.

Regeneration: The NaOH/C column was regenerated by soaking with 100 cc of water for 20 minutes. Water was drained and its chloride content determined. The column was soaked again with 100 cc of water. These steps were repeated until substantially all the chloride had been removed from the column. Four soakings removed 90 % of the chloride from the spent NaOH/C column.

Re uvenation: After regeneration to remove salt the column is blown dry with air. The dried carbon is impregnated in-situ twice with 50 % NaOH solution and dried with air. Since the regeneration did not remove all chloride the loadings in each subsequent rejuvenation became smaller and smaller.

The results are shown in Tables 1 - 4 and Figures 1 and 2.

F-7808

Table 1. HC| AdsoEBϋsni NaQHZAjeelism-vj Carbon

F-7808

F-7808

abe 3. Re ene e A /A c

F-7808

Table 4 Regenratjgn of Spent Adsorbent: NaQH/Centaur Carbon

Total ^'"

Solvent mg Cl/cc Solvent Total Cl NaOH,

Date Washβ _: volume.cc , Solvent used.BV recovered.mg pH %Recovery ; Cl, moles moles ■ Selectivity

Tall Glass Unit 1 w/Ceπtaur Carbon I I

33-17 85 3795 0.94282 J3.162E-06. 298145 2114 8.1 61.92 0.59549 1.259E-06 473017 1758 7.89 8185 0.49521 '7.762E-07| 637956 1015 7.65 93.36 0.28592 4.467E-07 640085

Discussion

The data show that NaOH/C is more reactive than both promoted alumina and NaOH bead in removing HCl from gases, as shown graphically in Fig. 1. HCl removal of over 99 % is achieved using a 40.6 cm column (table 1) . Apparently the NaOH is well spread over the high surface area of the activated carbon and became more readily accessible for reaction than NaOH in NaOH beads.

Essentially all of the NaOH on the activated carbon is accessible and reacted with HCl in the adsorption cycle, as shown if Figure 2. Stoichiometrically, the chloride loading is 0.89 g of Cl per g of NaOH. Figure 2 showed that about 0.9 g of Cl was loaded per g of NaOH. Thus the utilization rate of NaOH is about 100 %. The chloride removal efficiency did decrease somewhat as the NaOH was nearly exhausted and then fell sharply as the NaOH was exhausted. This indicates that the last trace of NaOH is situated at locations such as the end of a pore which are difficult to reach. The Atochem vapor phase carbon is a better support than Centaur carbon. The Atochem vapor phase carbon is lower in density, higher in NaOH loading capacity, high in HCl adsorption capacity and more reactive in HCl adsorption, as shown in Fig. 1. The chloride loading capacity of the NaOH/carbon is proportional to the NaOH loading on the carbon. The NaOH loading of NaOH on Centaur carbon was 18.1 % and its ultimate chloride loading capacity was about 19 % (Table 2) . Similarly, the NaOH loading on the Atochem vapor phase carbon was 40 % and its ultimate chloride loading capacity was about 38 %. Thus the carbon with high pore volume and NaOH loading capacity is a preferred support or base for use in preparing NaOH/C for HCl removal.

The spent NaOH/C adsorbent was effectively leached out by the water, as shown in Table 3. The effluent was a brine saturated with NaCl. The concentration of NaCl in

the regeneration procedure was successfully repeated four times.

As Table 3 shows, there is little NaOH in the effluent. The pH of the initial effluent was higher but never exceeded 9. Thus the utilization of NaOH loaded on the carbon for HCl reaction was essentially complete.

The regeneration procedure/rejuvenation procedure permitted re-impregnation of the bed in situ with NaOH. After drying with gas, the rejuvenated bed was ready for a second cycle of HCl adsorption. The regeneration/rejuvenation and adsorption cycle was successfully conducted for 4 times, as shown in Table 1 and Figure 2.

The regeneration operation was never complete. The NaOH loading of the first cycle was done outside of the unit for an initial NaOH loading of 40 %. The NaOH loadings after the 1st, the 2nd, the 3rd and the fourth regeneration were 30.1, 27.7, 24.3 and 20.2 wt %, respectively.

The loaded NaOH in each cycle was equally active for HCl removal. In terms of g of HCl adsorption per g of NaOH loaded, the HCl loadings were all around 0.9 (Figure 2) . In fact, the HCl loading increased somewhat as the cycle number increased. Apparently, the residual NaOH survived through the regeneration continues to react with HCl in the subsequent cycles.

These data and experiments show that the activated carbon can be reused again and again. Unlike the conventional alumina system there need be no disposal of spent adsorbent, it can simply be reused. Most refiners will prefer to make (or purchase) charcoal with 25 + wt % NaOH, regenerate it by water washing once or twice, and then discard it or send it out for re-processing. With the cost and potential liability which flow from waste disposal, many refiners will prefer to regenerate and rejuvenate their carbon beds, to reduce or eliminate

disposal of adsorbent. If disposal is used, the regenerated carbon is non-hazardous and can be disposed of easily.