4 Experimental part
4.4 Test run results
After each test, first evaluation of test results was visual appearance of the heater rod and heater tube assembly. In many cases the rod appearance was very clean, but still the outlet temperature showed significant descent from highest temperature during test. Good exam-ples of this observation were the tests conducted with higher flow velocity: the rods after test run appear as clean (Figure 23) as the initial clean rod (Figure 22).
Figure 22 – Clean heater rod
Figure 23 – Heater rods after 2.5h tests. From left to right CTO 2546, CTO 2546 repetition, CTO 2508, TOP 2137, TOP 2135
When processing the actual data, the test success was first evaluated from test run data charts (Appendix 1). From these charts it can be observed if the test has had some problems, for example fluctuating inlet temperature suggesting clogging of flow channel or pipelines, or if the equipment’s instruments had been interfered with surrounding devices magnetic fields.
In the past such interferences had been observed in test runs.
Two tests were ran with same crude tall oil sample CTO 2546 to see how similarly the sam-ple behaved when parameters were kept constant. In both tests the outlet temperature decrease was significant, the outlet temperature dedecrease being 35.1 °C in the first run and -45.3 °C in the repetition run as seen on Figure 24. Fouling resistance development was also similar and fast for both test runs as seen in Figure 25. Besides the significant decrease in
outlet temperature and high fouling resistance development, heater rods of both test runs appeared surprisingly clean as seen on Figure 23.
Figure 24 – Test outlet temperature drop with CTO samples, run time 2,5h, rod 330 °C
-50,0 -40,0 -30,0 -20,0 -10,0 0,0 10,0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Outlet temp drop [°C]
Run time [min]
Outlet temperature drop Flow speed 4 mm/s
R446 CTO 2546 R447 CTO 2508 R449 CTO 2546
Figure 25 - Fouling resistance development with CTO samples, run time 2,5h, rod 330 °C
With another CTO sample, CTO 2508, outlet temperature decrease from its maximum value was negligible (-0.3 °C). Outlet temperature fluctuated during the test between 233.3 °C and 234.5 °C, when the considered start point of the test was when outlet temperature reached 233.6 °C for the first time seen in Figure 24. Therefore, calculated fouling resistance ap-peared negative (-0.05) and it could be concluded that the tests run time was not long enough to generate any fouling seen in Figure 25. Also, rod appeared was very clean without any deposits (Figure 23).
Two test runs with two different TOP samples were ran with higher flow velocity. Results can be seen in Figure 26 and Figure 27. Rod appearance after both test runs was very clean as presented in Figure 23, even though the test with TOP 2135 showed significant outlet temperature decrease of -33.3 °C from its maximum value of 268.1 °C during the test.
-0,20
Figure 26 – Test outlet temperature drop with TOP samples, run time 2,5h, rod 330 °C
Figure 27 – Fouling resistance development with TOP samples, run time 2.5 h, rod 330 °C
-40,0
As above results show, 2.5 h run time was not long enough for fouling to start with some samples. Tests done with 2.5 h run time also suggested that complete clogging of test equip-ment is not likely with tested samples, even with longer test run times. To get more data about fouling resistance development and, on the other hand, get more deposition onto rods during the tests, it was decided to run these samples again with slower flow rate and hence longer run time (18 h).
Four different CTO samples were tested with 18 h test run campaign. Test run results can be seen on Figure 28 and Figure 29.
Figure 28 - Test outlet temperature drop with CTO samples, run time 18h, rod 330 °C
-65,0
Figure 29 - Fouling resistance development with CTO samples, run time 18 h, rod 330 °C
CTO 2508 showed only negligible decrease in the outlet temperature, even with the longer test run. As could be expected based on the shorter test runs, CTO 2546 showed significant decrease in the outlet temperature during the test. Maximum outlet temperature being 251.1
°C in the start, decreasing to 202.3 °C when test was run for 10 hours. After the test was ran for 930 min it was considered to become unstable (data fluctuating) and following results was cut out from the final results.
Three test series were conducted with lower rod temperatures and shorter test durations (300
°C run time 5 h, 270 °C run time 6 h and 240 °C runtime 12 h) in order to find out how fouling rate would develop in lower temperatures. Reducing maximum rod temperature clearly slows also overall fouling rate, as could be expected. After these test runs, the rods appeared very clean (Figure 30) compared to significant temperature drop (Figure 28), even when run times were shortened.
-5,00
Availability of CTO 2520 sample was limited, so the run time needed to be limited to 12h.
During this test CTO 2520 showed moderate decrease in outlet temperature (-12 °C), but when interpreting the shape of the fouling resistance plot (Figure 29) it can be assumed that fouling resistance was rather increasing than stabilizing. After the test, heater rod from CTO 2520 sample showed significantly different deposition formation compared to other samples presented in Figure 30.
CTO 2513 had only very minor reduction in outlet temperature -2.6 °C during the test and overall fouling resistance was almost as low as with CTO 2508.
Figure 30 - Heater rods after 18 h test with CTO. Test run details on top of the rod picture (sample ID, test target rod temperature, decrease in outlet temp.).
Images in Figure 30 are taken after letting the rod stand in a 70 °C oven for 1 h to remove excess liquid sample, yet a clearly visible oily film remains on top of the rods. Other rods,
except CTO 2520, had only slight colorization around rod hottest spot area (15 mm from top of rod), but no any clearly attached deposits could be visually observed. Solidified oily re-mains of feed had caused some (rosin) crystals to form on surface of rods CTO 2508 and CTO 2513 after cooling. After the test run with sample CTO 2546 the rod was covered with a particle rich gel-like layer seen in Figure 31. The gel-like layer was loosely attached to the rod. It slid loosely on top of the rod and fell off to the drip tray while dismantling the test setup. Similar, but less prominent, gel-like layers were observed also with other test runs, except processed ones and CTO 2508, when dismantling the setup.
Figure 31 – CTO 2546 rod and gel-like loose deposit on rod after test run (330 C, slow flow speed).
A sample of the gel-like layer was sent for further analysis from test runs with CTO 2546, CTO 2513, CTO 2520, TOP 2137 and TOP 2615.
Three different TOP samples were available for fouling tests. The results from the tests (in-cluding repetition tests) are presented in Figure 32 and Figure 33.
Figure 32 - Test outlet temperature drop with TOP samples, run time 18 h, rod 330 °C
Results with TOP2135 and TOP2137 samples were very repeatable, yet TOP 2135 showed more volatility in data during test. Fouling resistance development was much more similar among TOP samples than seen with CTO samples, as seen when comparing Figure 33 and Figure 29.
Figure 33 - Fouling resistance development with TOP samples, run time 18 h, rod 330 °C
Figure 34 - Heater rods after 18 h test with TOP samples. Test run details on top of rod picture (sample ID, test target rod temperature, decrease in outlet temp.).
0,00
Figure 34 presents the visual appearance of the heater rods after the test runs conducted with TOP samples. TOP 2615 rod appearance differs from the others because the gel-like deposits found during disassembling the heater assembly has been shovelled back on top of the rod to illustrate situation during the actual test. Gel-like layer typically slips away from the rod while disassembling the rod holder.
As the TOP 2137 sample demonstrated good repeatability and best overall stability during the test run and, on the other hand, caused clearly developing fouling resistance, it was se-lected to be tested with thiophene dosing described in chapter 3.3.3 (Stephenson et al., 2015).
Test was conducted with thiophene dosing of 0.5 w-% and effect of this dosing can be seen on Figure 35 and Figure 36. Needed dose of thiophene was weighed with analytical scale and added to feed reservoir after weighed amount of fresh feed was added to reservoir.
Figure 35 – TOP2137 outlet temperature drop with and without thiophene, rod 330 °C
-35,0
Based on this one test run with thiophene dosing, it seems that adding thiophene to the feed does not bring any significant difference on fouling rate. With thiophene dosing the fouling seems to start a bit delayed, but eventually it rose even higher than without the dosing. De-layed fouling resistance generation could be due to normal sample variations as the number of tests is too small to make conclusions from this minor difference. Based on this one test run, thiophene does not offer feasible solution to be used as a fouling inhibitor for tall oil based feeds at high temperatures.
Figure 36 – TOP2137 fouling resistance development with and without thiophene, rod 330
°C
As the foulant on top of the heater rod seen in tests were mainly loosely attached gel-like foulant layer, two test series were done with intermediate rinsing of the heater rod between
-2,00
the runs to see if heat transfer could be restored by a gentle flush with feed sample. The test run was split to three 5 h runs and the heater rod was flushed between the runs by dismantling the rod assembly and rinsing the rod gently with actual feed sample to see if gel-like layer could be removed with more turbulent flow conditions. Results of this variated runs can be seen on Figure 37 and Figure 38.
Figure 37 - Outlet temperature drop with intermediate rinsing of heater rod, rod 330 °C
Intermediate rinsing between the shorter runs returned heat transfer in both CTO and TOP tests close to clean state even though images taken between the runs shows slight coloriza-tion at proximity of hotspot area as seen in Figure 39.
-65,0
R474 CTO 2546 Interm. Rinse R451 TOP 2137
R462 TOP 2137
R480 TOP 2137 Interm. Rinse
Figure 38 - Fouling resistance development with intermediate rinsing of heater rod, rod 330
°C
Fouling resistance development is surprisingly similar after 5 and 10 h of test with the inter-mediate rinse compared to fouling resistance development with clean rod. The Figure 37 and Figure 38 suggest that the gel-like layer is actually contributing most to the overall fouling resistance development, rather than the visible colorization seen on the rod around the hotspot area after the tests.
0,00 5,00 10,00 15,00 20,00 25,00
0 60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960 1020 1080
Rf [m2 K /kW]
Run time [min]
Fouling resistance Flow speed 1 mm/s
R452 CTO 2546
R474 CTO 2546 Interm. Rinse R451 TOP 2137
R462 TOP 2137
R480 TOP 2137 Interm. Rinse
Figure 39 - Heater rods between intermediate runs, after rinsing with feed. From left to right CTO 2546 after 5 h, CTO 2546 after 10 h, TOP 2137 after 5 h, TOP 2137 after 10h run times
The overall appearance of the rods between the short runs seems very similar to the other runs (Figure 23, Figure 30 and Figure 34). During rinsing of TOP 2137 rod after 10 h run time, a small clean spot was made accidentally (seen in upper part of the rod on Figure 39).
Figure 40 - Test outlet temperature drop with processed samples and unprocessed samples, run time 18 h, rod 330 °C
Processed CTO and TOP samples was tested to see how different experimental processing methods developed in-house by Neste might effect on fouling rate. As seen in Figure 40 and Figure 41, processing of feed samples had a significant effect on the fouling rate. Neither of the processed samples of CTO 2546 showed any notable change in fouling resistance during the test runs. Processed TOP 2615 showed moderate development of fouling after 6 hours of test.
Figure 41 - Fouling resistance development with processed samples and unprocessed sam-ples, run time 18 h, rod 330 °C
Based on the performed test runs, it has become clear that fouling tendency varies greatly between the tested samples. Some of samples start to show signs of fouling immediately when stable state operation of test equipment is reached, when some of the samples do not show any signs of fouling with same conditions. Lowering the rod surface temperature slows the fouling rate, but even with as low as 240 °C test temperature fouling was rapidly devel-oping with the most problematic sample (CTO 2546).
However, deposition that contribute the most to the overall fouling resistance seems to be the gel-like layer that was loosely attached on top of the rod. The gel-like layer was identi-fiable after those test runs that showed significant outlet temperature decline during the test
-5,00
run. Processing of samples also eliminated (CTO 2546) or at least greatly delayed the devel-opment of gel-like layer (TOP 2615). Test duration was however only 18 h, therefore pro-cessed CTO samples might start to show signs on fouling significantly later than this. With processed TOP sample, fouling started much later than with un-processed sample and after 18 h test run fouling rate was rather accelerating than stabilizing, suggesting that longer test runs would probably show more clearly the true development on fouling resistance.
Table 11 – Summary of test results
Run Feed Feed ID
Summary of test results can be seen on Table 11. No clear correlation of fouling rate with any particular impurity or organic compound could be identified. When similar amount of in-organic impurities is found from feed sample, fouling rate of sample could be way differ-ent and hence predicting fouling tendency of tall oil based feeds seems impossible based on
impurity level. This can be concluded when comparing Table 11, Figure 20 and Figure 21.
However when overall impurity levels are reduced also fouling rate will reduce.