C 5 Vacuum Distillation
C 5 Vacuum Distillation
C 5 Vacuum Distillation
OGS
Page 1 of 20
Chapter 5 Contents
5.1 Introduction 5.2 Reduced Crude Flashing
5.2.1 Vacuum Bottoms Handling 5.2.2 Entrainment Control 5.2.3 Product Condensation 5.2.4 Vacuum Pressure Measurement
OGS
Page 2 of 20
Chapter 5
Vacuum Distillation
5.1 INTRODUCTION
In order to maximize the production of gas oil and lighter components from the bottoms material of an atmospheric distillation unit, these bottoms (reduced crude) can be further distilled in a vacuum distillation unit. Vacuum distillation of an oil means that the pressure on the oil being distilled is lower than the atmospheric pressure. It does not mean that there is a perfect vacuum above the liquid. The distillation of heavy oils is conducted at a low pressure in order to avoid thermal decomposition or cracking at high temperature. A stock which boils at 400 C at 50 mm. would not boil until about 500 C at atmospheric pressure, at which temperature most hydrocarbons crack. For distillation to take place, the vapor pressure of the liquid being distilled must be a little greater than the pressure above it. The molecules that comprise a liquid are held together by two forcesnatural cohesion and the weight of the atmosphere pressing down. This pressure is equal to 14.7 psi. at sea level and will support a column of water 34 ft. high. Now if boiling will begin when the vapor pressure of the liquid has become a little greater than the pressure holding down, it is clear that by removing some of the holding down force the liquid will start boiling at a lower temperature. The vacuum unit differs from the atmospheric type in that it has a fractionating column of larger diameter with bubble trays farther apart. This is necessary because much larger volumes of vapors have to be handled because of the lower pressure. Any sudden increase in vacuum will expand the volume of the vapor rapidly and possibly result in puking the tower. In the vacuum unit, almost no attempt is made to fractionate the products. It is only desired to vaporize the gas oil, remove the entrained pitch, and condense the liquid product as efficiently as possible. Bottoms from the crude tower contain material that can be charged to the catalytic cracking unit or be used for lube oil stocks. Distilling this material at atmospheric pressure would require high temperatures that would cause thermal cracking. Thermal cracking is undesirable because it would cause a loss of valuable product, degradation of valuable product, and shortened run time due to coke formation in pipes and vessels. For these reasons we conduct the distillation of the heavy reduced crude under vacuum in the vacuum tower.
OGS
Page 3 of 20
To achieve a deep vacuum, pressure drop through the column must be kept low. Instead of the type trays weve discussed earlier, we use random packing and demister pads, to keep the vapor velocities low a large diameter tower is used. An actual operating vacuum tower is show in Figure 5.1. The side draws from the vacuum tower may be lube oil stocks or charge to the Cat. Cracker. The bottoms, vacuum residual, may be heavy fuel oil or asphalt.
OGS
Page 4 of 20
Vertical Baffle
OGS
Page 5 of 20
OGS
Page 6 of 20
Some vacuum units are provided with facilities to strip the pitch with steam, and this will tend to lower the temperature necessary to meet an asphalt specification, but an excessive quantity of steam will overload the jets.
OGS
Page 7 of 20
Most of the gas oil spray is revaporized by the hot rising vapors and returned up the column. Some slop wax must be yielded in order to reject the captured entrainment. The amount of spray to the demister blanket is generally varied so that the yield of slop wax necessary to maintain the level in the slop wax pan is about 5% of the charge. If the carbon residue or the metals content of the heavy vacuum gas oil is high, a greater percentage of slop wax must be withdrawn or circulated. Variation in the color of the gas oil product is a valuable indication of the effectiveness of entrainment control. Slop wax is a mixture of gas oil and pitch, and it can be recirculated through the heater to the flash zone and reflashed, if the plant has the capacity to do so. If, however, crude contains volatile material compounds, these will be recycled with the slop wax and can finally rise into the gas oil. Where volatile metals are a problem, it is necessary either to yield slop wax as a product, or to make lighter asphalt, which will contain the metal compounds returned with the slop wax.
OGS
Page 8 of 20
The heavy vacuum gas oil circulation rate is chosen to maximize crude heat exchange. The best way of doing this on an operating unit is to observe the temperature of the crude leaving the crude / HVGO exchanger, then lower the HVGO circulation rate by 10%. If the crude temperature rises, the effect of the higher HVGO pan temperature has been greater than the effect reduced circulation, and we should try some more of the same % age. If the crude temperature is lower, we should try a 10 percent change in the opposite direction. Sometimes it is impossible to remove enough heat with crude exchange alone, and some HVGO from the outlet of the HVGO cooler is returned to the circulating line. This should only be done when necessary, since both heat and cooling water are wasted. The HVGO product is cooled and pumped to storage on HVGO pan level control. The LVGO section is a final contact condenser and normally the circulation rate should be adequate to keep the vapor to the jets within about 5C of cooling water temperature. A high circulation rate will provide a cushion against upsets.
OGS
Page 9 of 20
vacuum tower is swaged down to decrease the time that the bottoms liquid spends at the elevated bottoms temperature. A quench oil inlet line is also provided to protect the bottoms pumps. The feed line to a vacuum column is very large in comparison to the feed lines of most fractionators. This is because of the low pressure which causes almost all the vacuum column feed to be vapor. This also requires a special distributor called a tangential distributor that imparts a swirling direction to the feed and prevents damage to equipment above the distributor due to the rapid expansion of the feed as it enters the low pressure of the vacuum tower. The internals in the vacuum column are designed for a minimum amount of pressure drop, and the slop wax accumulator, the grid end demister pads are the only internals that extend completely across the entire column. The grid and demisters provide coalescing mediums to remove entrained liquid particles from the rapidly rising vapors. Spray distributors are used to aid the grid and demisters in coalescing. There are no trays in this vacuum tower, but what appear to be trays are side-by-side pans. Their outer edges perforated by holes and stiffened by metal lattice, the sideby-side pans overlap and provide a cascading effect to the condensed liquid. Hot vapors pass through the cascade to re-vaporize the lower boiling components of the liquid. Accumulator trays are designed to provide a vapor-free liquid to the suction of side draw pumps. Pump vents are returned to the column to allow removal of noncondensibles from the pump during startup. This helps a great deal in getting the pump started. After the pump is pumping properly, the vent should be closed. The top section of the vacuum column is swaged down because the traffic of material through the top of the column is much less than at the side draws. In fact, too many light ends in the feed or light ends formed by thermal decomposition of the bottoms would place an undue burden on the vacuum ejectors that have created and are maintaining the low pressure on the vacuum column. Vacuum columns are generally designed to withstand an internal pressure of 50 psi (3.5 Kg/cm2 gauge) and a 14.7 psia (760 mm absolute) external pressure. To strengthen the vessel walls to work between these two pressures, stiffeners are used. These are merely rings welded around the column and spaced a few feet apart.
OGS
Page 10 of 20
OGS
Page 11 of 20
The materials of construction used in the design of a vacuum tower are killed carbon steel for the vessel with the lower section clad with an 11-13% Cr. S.S. The slop wax accumulator is made of a 12% Cr. S.S. and the wall of the accumulator is lined with concrete. The grid is constructed of 304 stainless steel. The upper demister pad is constructed of Monel. Side-by-side pans are constructed of 12% Cr. S.S. The remainders are constructed of carbon steel. Designs for vacuum columns with different corrosion severities may allow the elimination of alloy cladding and some of the alloy in side-by-side pans and accumulators. Flash Towers Flash towers are used in services where good fractionation of the charge is not necessary. Their main purpose is to obtain the maximum amount of overhead product for a given transfer-line temperature. The feed material is heated to the desired temperature and the mixture of vapors and liquids is introduced into the flash space of a vessel. The liquid remaining after flashing falls to the bottom, usually through baffles to aid in separation and the vapors go overhead. Normally, an operation of this nature is used to remove asphalts and tar from crude before processing. Tar separators and tar stripper are the common names used for this type of tower.
pressure of the vacuum tower to atmospheric pressure. The steam used to pull the gases and is condensed in each stage and removed as water. Figure 5.3 is a schematic diagram of a vacuum jet system. Water is removed from the jet stages by a pump or gravity flow from a water column. If the jets are 34 ft. above ground level the water flows out by gravity. At any height lower than 34 ft. the water must be pulled off with a pump.
Figure 5.3 Schematic Diagram of Vacuum Jet System Barometric systems are generally controlled by changing the water flow to the condenser for the first-stage jet. Control can be attained by varying steam to the jet. This type of control is normally not as effective as varying the condenser water. Every vacuum system has a definite capacity. This capacity is measured as to the quantity of non-condensable removed at a definite vacuum. As the quantity of non-condensable exceeds the capacity of the jets, the vacuum begins to fail off. Therefore as the cooling water to the first stage jets is reduced the quantity of non-condensed gases exceeds the capacity of the jet and causes the vacuum to fail off.
OGS
Page 13 of 20
stage discharge header is, of course, higher than the pressure in the suction header, but if the velocity of the steam through the diffuser throat is high enough, gas cannot back into the suction header. If a single ejector is incapable of raising the gases to atmospheric pressure at which they can be vented, the steam is condensed and a second ejector taking suction of the non-condensable gases raises them to a higher pressure. The dimensions of an ejector are quite critical so that any given ejector will only operate ever a relatively limited range. A substantial change in suction or discharge conditions will probably demand a change in the dimensions of the nozzle or diffuser, or both. The effect of changes in operating conditions can be summarized as follows: 5.4.2.1 Steam Pressure Must be maintained quite close to that for which the equipment was designed. If the steam pressure greatly exceeds that for which the nozzle was designed, the quantity of steam discharging into the diffuser will be greater than can pass through the diffuser, and steam will back into the suction header. Too low a steam pressure will mean a drastic loss in performance of the ejector. It should be noted that wet steam will cause random fluctuations in ejector performance and in addition will erode the nozzle and diffuser. 5.4.2.2 Discharge Pressure If the discharge pressure rises above design, there is an increasing probability of reverse flow. An increase in discharge pressure on an ejector discharging to atmosphere is only possible if the discharge is obstructed. But on multi-stage units an increase in interstage pressure due to high condensate temperatures or failure of a second or third-stage ejector will immediately affect the performance of the firststage unit. 5.4.2.3 Load A decrease in load (kg/hour of vapor to ejector) will result in a somewhat higher vacuum being obtained, but if the load is increased above design, the vacuum obtained will fall off quite suddenly and dramatically.
OGS
Page 14 of 20
OGS
Page 15 of 20
OGS
Page 16 of 20
5.4.2.4 Cooling Water Temperature The temperature at which the steam is condensed in the inter and final condensers will have a relatively minor effect on the vacuum obtained, but will substantially reduce the load at which the ejector system breaks down, since an increase in condensate temperature increases the inter-stage pressure. In order to insure flexibility a refinery ejector system for a vacuum unit will generally be constructed using two parallel sets. The minimum combination of equipment which will achieve a satisfactory vacuum is normally used. The vapors drawn from the top of a typical vacuum unit to the jets consists of air from leaks, steam entrained from the bottom of the crude distillation tower, light hydrocarbons, and sulfur and nitrogen compounds from thermal decomposition in the heater, and any hydrocarbons lighter than gasoline which have not been stripped from the charge. The steam and light hydrocarbons will condense in the inter-condenser so that the first-stage ejectors can be heavily loaded under conditions which only lightly load the second-stage ejectors. Any actual cracking in the furnace will produce light gases which will very rapidly overload the second-stage ejectors. Both the condensable and non-condensable vapors handled by a typical vacuum unit ejector set are highly odiferous so that the condensate must be stripped and the noncondensable vapors incinerated.
OGS
Page 17 of 20
5.6 STARTUP
When starting a vacuum unit it is common practice to steam out the heater and tower and then pressure test the tower with steam at about 50 psig. It will be assumed that this has been done and that all valves around the ejectors are closed. a. Shut off steam to the tower. b. Vent steam from top of tower until the pressure is about 0.2 Kg/cm2, then close end plug the top vent. c. Open the inlet and outlet valves of one of the second-stage ejectors. d. Open the inlet and outlet valves of one of the first-stage ejectors. e. Open water through both condensers. f. Open steam to both first and second-stage ejectors. g. Check that the steam is dry and adjust the steam pressure to that given on the equipment name plate. As soon as a level appears in the inter-condenser, start the condensate pump and place the level on control.
OGS
Page 18 of 20
h. Let the ejectors run until a constant vacuum is obtained even though the presence of water in the tower may result in a poor initial vacuum. i. Charge the vacuum unit and proceed with normal operations. Adding Additional Ejectors The operators should observe and get to know the interstage pressure which gives the most stable operation on a given ejector set. On most two-stage units, this is about 260-130 mm Hg. absolute vacuum. When the tower vacuum either decreases or becomes sensitive to process conditions, additional ejector capacity should be added. If the interstage pressure has risen (the vacuum as measured in mm Hg. has increased), an additional second-stage jet should be added. If the interstage pressure is unchanged, but the lower pressure has risen, an additional first-stage jet should be added and this may render the addition of another second-stage jet necessary. a. Open the ejector discharge valve. b. Open the steam to the nozzle. c. Check that the nozzle steam pressure is under good control. d. Open the ejector suction valve.
5.7 Troubleshooting
Occasionally, ejectors will fail to pull an adequate vacuum or will perform erratically. This can be the result of a large number of troubles, a few of which are listed below for checking. a. Air leaking Into the System: Hot bolt all flanges and manways on the vacuum tower and on the overhead line. Tighten and oil all valve packing glands. Plug all vents and drain valves and tighten any screwed connection. Check that pump gland flushing is adjusted to maintain a positive pressure on the gland. b. Air Leaking Into Interstage: This will be confirmed by a rise in interstage pressure. Tighten all flanges, pump glands and screwed connections. Check that the condensate drain trap is not stuck and bleeding air back from the second stage.
OGS
Page 19 of 20
c. Leaking Behind Nozzle: Certain styles of ejectors can readily leak air or steam through a leak in the point where the nozzle is attached to the body. d. Wrong Steam Pressure: Check ejector name plate data and change pressure controller setting. e. Wet Steam: Causes erratic performance. Check performance of steam traps. f. Worn Nozzles and Diffusers: The result of using wet steam. g. Clogged Steam Filters: There is generally a main filter ahead of the steam pressure controller and a filter in the nozzle of each ejector. h. High Condensate Temperatures: The result of insufficient cooling water flow or fouling of either the tube or the shell side of the condensers. i. Flooded Condensers: The result of malfunctions of the level controller or the condensate drain trap, or of pump failure. If the pump will not hold suction, check that air is not leaking in at the gland. l. Faulty Installation: Failure to properly align gaskets and similar details which are normally insignificant can trap condensate pockets or cause turbulence which can affect the performance of vacuum equipment. If an ejector set has been dismantled, each nozzle must of course be reinstalled in the correct diffuser. j. Back Pressure: Due to deposits in the condensers, a plugged flame arrester in the vent, a condensate pocket or other obstruction. k. Impossible Operation: Such as attempting to obtain an absolute pressure lower than the vapor pressure of a liquid in the system. If a very little vacuum gas oil is produced it may be so light that it will be impossible to pull a vacuum on the tower. m. High-Tower Bottom Level: If the level in the tower is permitted to rise, some cracking will occur because the pitch is being held at a high temperature for too long. n. Steam Entering System: Check steam-out connections on the tower, heater, exchangers, etc. o. Cracking in Furnace: Experienced at very high transfer temperatures and can be checked by running an Oliensis on the pitch, or a bromine number on the light vacuum gas oil. Where a leak is elusive, a special meter can be installed to measure the noncondensable being vented and these vent gases can be analyzed in the laboratory.
OGS (Rev.0) Feb. 2008 Page 20 of 20