Originally appeared in: July 2009 issue, pgs 35-39
International PETROLEUM REFINING Reprinted with publisher’s permission.
A New Approach to
Improving Heater Efficiency
By Ashutosh Garg, FIS has developed a new approach to
improve heater efficiency that cuts
Fired heaters are important equipment in any Re-
finery or Chemical plant. In the USA , we have down the cost of revamp by ½ to 1/3
about 155 refineries operating today. The esti-
mated number of heaters in the refining industry
is about 3200. A major refinery may have about and improves the payout significantly.
20 to 40 heaters depending upon the size and
complexity. There are about 1400 heaters in the
Chemical Industry. Half of these heaters are in
high temperature pyrolysis service for making eth- ant section and produces a temperature of about
ylene and propylene. About a quarter of these 3200° F. The flue gas temperature leaving the
heaters are in Steam Reforming of hydrocarbon. radiant section is 1650° F. About 2/3 of the heat
All the major industries are highly energy inten- duty is absorbed in the radiant section and the
sive. rest is absorbed in the convection section.
A typical refinery of 100,000 BPD size In the convection section, flue gases are
would be consuming almost 1.84 Billion Btu/hr and cooled to 750° F . The estimated thermal effi-
the fuel bill is close to 100 Million Dollars. Even a 1 ciency of this heater is about 80%.
% efficiency improvement will add about a million If we want to improve the efficiency of the
dollars to the bottom line of the heater. It is our fired heaters in a conventional way, then the
experience that about 1-2% efficiency improve- most common way is to reduce the flue gas tem-
ment can be accomplished with just good house perature approach to the inlet fluid temperature.
keeping without any major capital investment. In Figure 2 we are showing that an additional
Typically in a fired heater, the feed is pre- convection section is installed in series to the ex-
heated in the convection section and is further isting convection section. This is fine for most of
heated in the radiant section. Figure 1 shows a the applications but it does have a drawback. The
typical 4 pass heater with feed being heated from pressure drop on the fluid side goes up. In a num-
450 to 600° F. Fuel gas or oil is burned in the radi-
Flue gas 500 °F Flue gas 750 °F Flue gas 1,650 °F
Flue gas 750 °F Flue gas 1,650 °F
500 °F Feed Out Additional 500 °F Feed Out
Convection Section Radiant Section Convection Convection Section Radiant Section
450 °F 600 °F
Section 600 °F
Flue gas 1,650 °F Flue gas 3,200 °F
Flue gas 750 °F Flue gas 1,650 °F Flue gas 3,200 °F
Figure 1.Typical 4 pass heater with feed being Figure 2. Conventional scheme for improve the
heated from 450 to 600° F efficiency
PRODUCTIVITY IMPROVEMENT DESIGN MODIFICATION
Flue gas 500 °F
pressure drop is a very important parameter and
Split Flow in
Split Flow out
lower pressure drop helps save power on recycle
Section- II 600 °F gas compressor as well as improve the yield of the
Flue gas unit. In these heaters, fluid is heated mostly in the
radiant section. The radiant coil consists of a num-
Flue gas 750 °F Flue gas 1,650 °F ber of parallel passes leading to the low pressure
drop. Convection section consists of waste heat
recovery which could be hydrocarbon preheating
Convection Feed Out
450 °F Section - I Radiant Section
or steam generation.
Our patented split flow process is very
suitable for this service as the feed can be easily
Flue gas 1,650 °F Flue gas 3,200 °F
split and sent to the radiant and convection sec-
tion at the same time (Figure 4).
Figure 3. FIS Split Flow* scheme to improve the
efficiency of the heater *Patented Case study
A Refinery had a reformer heater in their facility.
ber of plants, it is not possible to increase the In Jan. 2000, the fuel gas prices touched $10 per
pressure drop across the heater as that will re- MSCF. The client started a project to improve effi-
quire upgrading the pump and motor, may also ciency of the heater. The stack temperature in this
need to change the piping class. heater was 1100° F. The refinery did not want to
Our patented scheme called Split Flow generate steam as they have enough in their facil-
(Figure 3) divides the flow entering the heater into ity. They did not want an air preheating system
two streams- the main stream and the split due to side fired burners and complications associ-
stream. We send about 10-20% of the flow to the ated with them. They had tried installing air pre-
split stream and heat it to the outlet temperature. heater about 20 years ago during the first energy
The main stream continues to go the existing con- crunch but did not install the system as it became
vection section and radiant coils. The major ad- complicated.
vantage of this scheme is the reduction in pres- The heater duty was 158 MMBtu/hr out of
sure drop. As compared to the series flow which which 120 MMBtu/hr were used to heat the proc-
increases the pressure drop across the heater, in ess streams in the radiant section. The convection
split flow we see a reduction in pressure drop section consisted of splitter reboiler service and
across the heater coil. This offers a significant ad- stabilizer reboiler service. The firing rate was
vantage in pressure drop limited heaters or the around 234 MMBtu/hr and the thermal efficiency
heaters that are already running at the maximum was around 67% (Table 1).
charge rate. In current times, it is fairly common The heater was a bottleneck in the plant
to have heaters operating at maximum capacity. and was limiting the capacity. The client could not
One of the major building blocks in refin- fire it any harder because it would exceed the per-
eries is the catalytic reforming unit, which is used mitted firing limit.
to upgrade the octane number of gasoline. It is In the scheme of the Reformer heater be-
known by different trade names as Rheniforming, fore the revamp, the radiant section consisted of
Powerforming, Platforming etc. In this unit, the four radiant cells separated by bridgewalls. The
burners were installed on the end walls. The radi-
ant tubes were U shaped in the first two cells.
Each process stream had a large 26 inch inlet and
outlet manifold. There were 44 parallel coils or
tubes in each coil. Feed was entering the heater
Flow in 1,100°F
at 840° F and was heated to 1010° F in the radi-
ant section. The Reboiler process streams recov-
750°F Split Flow Out
Process Feed Out
Process Feed In 1000°F Process Feed In Process Feed Out
Parameter Units Operating
Total Heater Duty MMBtu/hr 158
Radiant Heat Duty MMBtu/hr 120.09
Convection Heat MMBtu/hr 37.91
Figure 4. Scheme of a reformer heater with
Firing Rate MMBtu/hr 234
(right) and without (left) the Split Flow technology.
Efficiency % 67.5
DESIGN MODIFICATION PRODUCTIVITY IMPROVEMENT
Stack Stack Stack
In #4 In
DAMPER #1 In #2 In Out
#1 Out #2 Out
INLET OUTLET INLET OUTLET INLET OUTLET
Figure 7. FIS Split Flow Scheme. Patented tech-
Figure 5. Reformer heater before the revamp. sure drop across went up by 50%.
The conventional design is shown in Fig-
ered heat from flue gases in the three convection ure 6. In order to keep the pressure drop low, the
sections. They were entering the convection sec- number of parallel passes had to be increased in
tion at 428° F and 459° F respectively. the convection section leading to very wide con-
The existing heater consists of 3 separate vection sections. Since all the feed had to be
convection sections and 3 stacks (Figure 5). This heated in series , it required large pipe sizes. The
heater design is very old and new heaters have pipe design becomes difficult as the temperatures
single convection sections. The convection sec- are high and thermal expansion needs to be taken
tions were of different size, first one was 4 tubes care of. The heater foundation could not sustain
wide, second was 6 tubes wide and the third one the new convection section loads. A grade
was only 2 tubes wide. Radiant section was di- mounted stack had to be installed.
vided by bridge walls. The stack dampers in the The conventional efficiency improvement
stack were not working well. scheme suffered from the following disadvantages.
The conventional design would be to pre- The process side pressure drop went up. The large
heat the process fluid in the convection section piping size was required to keep pressure drop
first and then send it to the radiant section. In low. Due to the large convection sections and
this design, the heat is recovered by the feed in grade mounted stack, the cost of the scheme was
the convection section and then by the reboiler estimated at 6 million Dollars.
coils. This would have improved the efficiency to Split flow technique divides the process
the desired level of 86%. fluid into two streams (Figure 7). The convection
The major disadvantage was an increase section heats approximately 25 – 30% of the total
in pressure drop across each coil. Due to inherent flow and the rest is heated by the radiant section
construction of the convection section with multi- to the same temperature. The exiting fluids from
ple return bends in the convection coil, the pres- both sections are then combined into one stream.
In the split flow design for this reformer
INLET INLET OUTLET INLET OUTLET
Figure 8. Revamp of the reformer heater with
Figure 6. Conventional design for efficiency im-
the Split Flow technology.
PRODUCTIVITY IMPROVEMENT DESIGN MODIFICATION
Table 2 Table 3
Original Split flow Before After
Design Design Revamp Revamp
Pressure Drop, psi 3.1 2.1 Capacity, BPD 18,500 24,000
Firebox temperature, °F 1,615 1,551 Heat Duty, MMBtu/hr 158 194.5
Radiant flux, Btu/hr ft² 19,823 15,047 Heat Release, MMBtu/hr 234 225
Radiant tube metal temp, °F 1,151 1,120 Efficiency, % 67.5 86.6
Firing rate, MMBtu/hr 116.35 82.65 Stack Temp., °F 1,092 478
heater (Figure 8), the convection sections are nar- Fuel Savings, $/annum 5.8 Million*
rower and lighter and as a result could be sup- *Based on $6.0 / MM Btu
ported on the existing heater structure with minor to 15000 Btu/hr ft2. The tube metal temperature
reinforcements. The split flow piping is smaller. went down by 30° F. The firing rate went down by
We used only 16-18 inch pipe size as compared to 33%.
26 inch pipe in the conventional scheme. The Table 3 compares some of the most im-
stacks could be reused after the new dampers portant parameters in the heaters operation be-
were installed on the stacks. The flow to the split fore and after the revamp. As you can see, with
coil was controlled using restriction orifices in each the split flow technology, the heater was able to
split flow line. This way the pressure drop was process 24,000 BPD of feed, up from 18,500 BPD.
balanced. The heat duty was up by 20% and yet the firing
Table 2 compares the performance of the stayed the same. Efficiency of the heater was in-
heater before and after the revamp. This data creased by almost 20%. The stack temperature
shows that the pressure drop across the cell 1 was reduced by 600° F. The project payout in less
went down from 3.1 to 2.1 psi. It would have than 6 months.
gone up by 50% in the conventional scheme. The Figure 9 shows the reformer before re-
reduction in pressure drop was a significant bene- vamp and after revamp. The Split flow technique
fit to the client for yield improvement. The firebox provides efficient solutions to pressure drop and
temperature was reduced from 1615 to 1551° F. capacity problems. It provides an inexpensive effi-
The radiant flux decreased from 20000 Btu/hr ft 2 cient alternative with a short payback time.
the FIS Split
Ashutosh Garg is currently working as a Thermal Engineer at Furnaces Improvements in Sugar Land, TX. He has more than
35 years of experience in design, engineering and troubleshooting of furnaces and combustion systems for the refining and
petrochemical industries. He graduated from Indian Institute of Technology, Kanpur, India, in chemical engineering in 1974.
He started as a graduate engineer in an ammonia plant. This was followed by six years in KTI India and eight years at EIL,
New Delhi, in the heater group. He joined KTI Corp., San Dimas, California, in 1990, and moved to Houston in 1992. He has
published several articles on fired heaters and burners in trade magazines. He is a registered professional engineer and a
member of AIChE, API & ASME.
4 *Reproduced with the permission of International PETROLEUM REFINING.