Question 34: Fresh sulfuric acid fed to an alkylation unit can contain niter (nitrosylsulfuric acid) which may lead to excessive corrosion. What is niter; what does it do; can we test for it; and how can we reduce the levels in our fresh acid?
MUEHLBAUER (Valero Energy Corporation – Benicia Refinery)
“Niter” is actually a common term referring to the amount of NOx (nitrogen oxide) or nitrates in the sulfuric acid. It is actually generated in the sulfuric acid regeneration process. One of the first steps you go through in acid regeneration is a combustion furnace. In that combustion furnace, you can generate NOx, which is really a function of your peak flame temperature and excess oxygen: the normal NOx contributors. But then, what the regeneration facility does with
that NOx is what can cause a variation in the niter concentrations. Some units are designed with a mist break eliminator that essentially can entrain this niter in the product fresh sulfuric acid, in which case the concentration in that stream would be much higher.
Other facilities either employ different technology for NOx control, or they are able to segregate the niter and reroute it to the spent acid. The niter stream becomes a feed to the regeneration unit, which would then get reprocessed and not end up in the product fresh acid. Within Valero, we actually have not had much experience with niter-related corrosion, nor have we had any of our corrosion caused by niter. Our acid supplier told me that our niter concentration typically ranges from 25 ppm to 50 ppm levels.
PIZZINI (Phillips 66)
I want to give a nod to Randy Peterson who provided a lot of the information for this response. As Joe said, niter is the NOx compound that can exist in the finished sulfuric acid. There are wet chemistry methods that can quantitatively measure how much niter is present. One method is a color change test with iron sulfate. I know that DuPont and Rhodia both have methods. You need to get in touch with either of those companies to get the method. The amount of NOx is influenced by how hot the acid regen furnace is operating. So, if you are burning just all-natural gas for heat, this high temperature flame will increase the NOx versus, say, a spent acid furnace that is burning acid gas. We have input indicating that the typical niter levels from a two-stage spent acid furnace are 25 ppm to 40 ppm.
The chart on this slide was created based on an equation included in an article in a 1980 metallurgical trade journal. I plotted out the information to show the corrosion rate of carbon steel in mils per year, based on a scale of 0 to 300, relative to how much nitrate is present in strong sulfuric acid at 122ºF. So, if you get upwards of 800 ppm of nitrates, corrosion could be very aggressive. The good news is that the NOx is consumed in that reaction, so it is not self0-sustaining and does eventually disappear in a closed system. If you get above 1,000 ppm of niter in acid, then the corrosion rate will actually drop as a result of having a passivation effect.
PIZZINI (Phillips 66)
“Niter” is a term that describes the amount of dissolved nitrogen oxides in sulfuric acid. The term is applied to the amount of either NO3 (nitrate) or NO2- (nitrite) present. This is a relevant question, because increased levels of niter will affect the corrosion rate of the acid on carbon steel since these compounds attack the protective ferrous sulfate coating on the metal surface. There are “wet-chemistry” methods to determine the amount of niter in sulfuric acid, available from DuPont and Rhodia. This test uses ferrous sulfate which reacts with nitrates and nitrites to produce a red color. The amount of niter in sulfuric acid is related to the conditions in the spent acid regeneration furnace, where hotter flames (e.g., burning only natural gas) or the presence of ammonia in the fuel gas will lead to higher NOx levels in the resulting fresh acid. For example, a two-stage spent acid furnace would be expected to have niter levels in the 25 ppm to 40 ppm range in the fresh acid.
Corrosion rate effects on mild carbon steel in concentrated sulfuric acid were documented in a 1980 AIME (American Institute of Mechanical Engineers) trade magazine, Metallurgical Transactions A, Volume 11A. The chart below was creating using information from that article. Note that the corrosion rate increases with nitrate concentrations up to approximately 1,000 ppm, and then decreased markedly at higher nitrate concentrations, where the steel passivity is re-established. Also note that the nitrates and nitrites are reduced (converts to NO and N2O gasses) as they accelerate corrosion and are therefore self-limiting in closed systems.
RANDY PETERSON (STRATCO® - DuPont)
We have recently seen excessive corrosion problems in sulfuric acid alkylation units that are being run very close to optimum design conditions. After significant trouble shooting and testing for many contaminants, we believe that unit corrosion is significantly accelerated by a contaminant found in all fresh sulfuric acid. This contaminant is nicknamed “niter” and is mostly composed of nitrosylsulfuric acid (NSA). Its chemical formula is NOHSO4 and it acts as an
oxidizer that apparently is very corrosive to carbon steel and Alloy 20. Niter is formed in the sulfuric acid regeneration plant (SAR) and is most concentrated in the drains from the mist eliminators within the absorption towers (i.e., final absorption and interpass absorption towers). These drains can either be allowed to drip into the tower bottoms where the niter combines with the product acid or can be collected within the towers and routed to another destination. This destination is typically either the 99 wt.% product acid tanks, the 90 wt% spent acid tank (SAR feed) or directly into the SAR furnace.
We have found that if the niter-rich drains are sent to the product acid, the niter content in the fresh acid sent to the alkylation unit can be as high as 300 ppm. If the drains are sent to the spent acid tank or directly to the furnace, most of the niter is converted to N2 within the SAR and harmlessly exits the stack with the flue gas. The product acid then typically contains less than 30 ppm niter. We suggest that you contact your sulfuric acid supplier and ask them, “Where are you sending the niter-rich drains from the mist eliminators?” Of course, we recommend that they not be added to the fresh acid that you purchase. If you have your own onsite SAR and the drains are not routed to the furnace or spent acid tank, we recommend that you contact your SAR licensor for Best Practice recommendations. If you would like DuPont’s colorimetric test procedure for measuring niter in fresh acid, please email me at: j-randall.peterson@dupont.com.
Here are a few pictures of what we believe to be niter-induced corrosion. The first picture is of Alloy 20 Acid Wash coalescing media that had gone “active” and collapsed within two years. This is typically considered a very mild service, liquid-full, with mostly isobutane, some alkylate, and less than 5 LV% of 99 wt.% acid. The temperature is approximately 85°F and the velocity is extremely low (<10 ft/min (feet per minute). The Chevron-type Alloy 20 coalescing
media originally filled the entire area between the wire grids. Notice that the carbon steel vessel and the 316 SS (stainless steel) perforated plate and support grids appear to still be in very good condition. Alloy 20 is a specialty metal that is specifically formulated for harsh sulfuric acid services. The failed Alloy 20 was analyzed by DuPont metallurgists and found to have the proper composition.
The following pictures are samples cut from a prematurely failed carbon steel Contactor™ reactor tube bundle. These tubes failed in less than two years under what is typically considered mild conditions (<50°F and 92 wt% acid). Pitting is not a typical failure mechanism of carbon steel in sulfuric acid service. This is the primary clue that led us to look for the presence of oxidizers such as niter.
Question 32: What are the impacts of the presence of acetone in the alkylation unit feed? How is this formed in the FCC? Comment on both HF and sulfuric units.
STEVES (Norton Engineering Consultants, Inc.)
In both HF and sulfuric acid units, acetone will consume acid, resulting in a reduced acid strength. Regarding sulfuric acid units, some documentation that I got from DuPont STRATCO indicated that one pound of acetone will consume about 10.5 pounds of acid. When talking with some HF alkylation experts, I was told that one pound of acetone in an HF unit will consume about half a pound of acid.
Acetone formation in FCC is suspected to be caused by oxygen carry under from the regenerator into the reactor section. I do not have a lot more detail on that. I would suggest that maybe you ask the FCC panel and see if you can stump them.
Typical acetone concentrations in the alkylation unit feed are about 100 ppm (parts per million) to 200 ppm. In HF alkylation units, levels above 250 ppm can lead to significant production of light ASO (acid-soluble oils), and the acid can become deep red as the acid strength falls.
MUEHLBAUER (Valero Energy Corporation – Benicia Refinery)
Similar to Chris, my experience is that acetone in the alkylation unit feed makes water; so, it requires more acid regeneration for the HF units and higher acid makeup rates for sulfuric units. Within Valero, we do not see acetone as a primary contaminant for our alkylation units. There is only one of our refineries that we actually test for acetone. Most of the oxygenates we found from the FCC were actually phenolic, so they would boil more in the gasoline range rather than end up in the alkylation unit feed.
I want to make one comment. In refineries that have configuration with an upstream oxygenate unit like MTBE/TAME (methyl tertiary butyl ether)/ (tertiary amyl methyl ether) or an isooctene unit, if there is poor fractionation in those units, you could get carryover of alcohol derivatives, such as acetone and dimethyl ether, which would impact alkylation feed. It is obviously becoming much less common in the U.S.; but for international facilities, it may still apply.
DAVID SMITH (UOP, A Honeywell Company)
A potential solution for removing the acetone from the alkylation unit feed is to use the hybrid adsorbent AZ-300. This adsorbent removes a wide variety of oxygenates, including acetone, as well as the water from the alkylation feed.
KURT DETRICK (UOP, A Honeywell Company)
Actually, the acetone can make water in an HF alkylation unit, but not all of it does. In fact, 100 ppm to 200 ppm acetone in the feed to an alkylation unit is not unusual, although it is usually a little less. If that much water actually came into the alkylation unit, you would know it. It would be pretty bad. So, we believe that some of the acetone probably makes water, but not nearly all of it. The majority of it probably works its way out the bottom of the regenerator in most units. At the typical levels that were listed, I think that most units are able to push it out the bottom of the regenerator over time. It does build up in the acid a bit. We have actually done sampling and found acetone in the circulating acid, so it is somewhat stable. One possible cause of high levels of acetone in the FCC unit LPG (liquefied petroleum gas) is insufficient deaeration of the steam used in the FCC stripping section. This allows oxygen to get into the reactor and provides the possibility of making acetone.
RANDY PETERSON (STRATCO® - DuPont)
We have analyzed olefin feed samples from a refiner that complained of higher-than-expected acid consumption in a sulfuric acid alkylation unit. During this analysis, we were surprised to find high levels of acetone ranging between 700 wppm (weight parts per million) and 1200 wppm in the butylene feed from an FCC.
Acetone consumes acid at a rate of 10.6 pounds of acid per pound of acetone [99.2 wt% (weight percent) to 90.0 wt% acid spending range]. Besides consuming significant acid, this contaminant increases the water content (relative to red oils) in the acid, which causes higher corrosion rates within the alkylation unit.
We are not experts on FCC operation, but we have been told that excess instrument air to the FCC reactor may contribute to high acetone levels in the alkylation unit feed.
Question 33: Increased feed sulfur increases acid consumption. How does it affect alkylate yield and/or alkylate properties?
MUEHLBAUER (Valero Energy Corporation – Benicia Refinery)
In HF units, when sulfur is in the feed, it produces acid-soluble oil (ASO), organic fluorides, and polymers, which then have to be removed through the regeneration process. The light ends that are contained in this ASO can put more pressure on the regeneration system and lead to the higher acid losses. So that is the mechanism there. But within Valero, we found that higher ASO really has a minimal impact on alkylate yields. We do see that it can impact the
product sulfur concentration. In one of our facilities, we run as high as 20 ppm sulfur in the alkylate. The main reason for this is ASO entrainment.
We have seen that units which practice internal regeneration are more susceptible to this entrainment, and that is the main area it would come out. We have also found that in units designed with vertical settlers, it is really important to monitor the superficial velocity and acid quality in those separators to make sure you are getting adequate separation.
One other point, which is not on the slide, is that we do have one facility with a sulfuric acid alkylation unit that runs as high as 500 ppm to 1,000 ppm sulfur in the feed. At those levels, we have seen high acid consumption but have not really seen the impact on the distillation properties of the alkylate. In fact, the product sulfur in that alkylate is really low: less than 3ppm. We believe part of that is because of the neutralization section and acid wash that is performed in that unit.
STEVES (Norton Engineering Consultants, Inc.)
I will just echo what Joe said. With normal feed sulfur levels of 10 ppm to 20 ppm maximum, the impact on alkylate yields and properties is minimal. With higher feed sulfur, the acid consumption will become more pronounced and can eventually result in a loss of alkylate yield as you make undesirable acid species.
In HF units, severe sulfur contamination and feed can result in a lot of light ASO production and eventually a yellow discoloration of the alkylate, especially, as Joe said, with internal regeneration. Alkylate endpoint could also increase an octane drop. In sulfuric acid units, severe and prolonged sulfur contamination can lead to an acid runaway, in which case the alkylate yield would drop in an acid runaway situation. Octane would decrease, and the alkylate can turn purple.
KURT DETRICK (UOP, A Honeywell Company)
Just one more comment on the alkylate properties. I think that the sulfur itself is not directly affecting the alkylate properties, such as the endpoint or the color. Rather, it is the drop in acid purity, which occurs as a result of a sulfur upset, that tends to cause these high endpoints and the color in the alkylate. The reason that happens is because the ASO that comes from the sulfur tends to be a fairly light boiling ASO. So, if you try to run the regenerator or rerun column
at about the same temperatures as normal, then you might boil it overhead. It stays in the system that way instead of getting rejected in the bottom of the rerun column or regenerator. If you know you had the sulfur upset and can drop the rerun or regenerator temperatures a little ahead of time to get that ASO out the bottom of the regenerator, then the effect on the alkylate endpoint will be minimized.
RANDY PETERSON (STRATCO® - DuPont)
We were contacted by a refiner that has very high levels of sulfur compounds in the olefin fed to their sulfuric acid alkylation unit. Most butylene feeds are treated with caustic washes or mercaptan extraction units prior to alkylation and usually contains less than 20 ppm total sulfur. However, this particular refiner’s feed typically contains approximately 500 ppm and sometimes well over 1,000 ppm of sulfur compounds. They see a strong correlation between the feed sulfur content and acid consumption which we would expect. However, they see much higher acid consumption than what we predicted.
We performed a series of pilot plant runs where we spiked the feed with various quantities of typical sulfur contaminants, such as ethyl mercaptan. In summary, we found that the sulfur components have about double the acid consumption than what we have previously published. For example, ethyl mercaptan consumes about 31 pounds of acid per pound of contaminant (99.2 wt.% to 90.0 wt.% acid spending range) versus the 15.7 pounds of acid per
pound of contaminant that STRATCO® DuPont has published in the past.
As far as alkylate product specifications, we saw no change in D-86 T90 or endpoint with changes in feed sulfur in our pilot plant. The refiner also reported that the alkylate product typically contains less than 10 ppm sulfur no matter the quantity in the feed. They have a modern effluent treating system with an Acid Wash which may help explain the low sulfur levels in their product.