Smart design, planning, execution can slash ULSD revamp costs dramatically

Diesel Fuel News, May 12, 2003 by Jack Peckham

German refiner Bayernoil found that it could slash the cost of a recent ultra-low sulfur diesel (ULSD) hydrotreating revamp project to as low as $106 per barrel per stream day, defying conventional wisdom that similar revamps must cost several times as much.

In a paper presented to American Institute of Chemical Engineers last month, Bayernoil's Klaus-Dieter Rost and New Jersey-based hydroprocessing consulting engineer Tai-Sheng Chou (of H2-Advance) showed how costs to produce ULSD (less than l 0 ppm sulfur) can be cut dramatically.

Several factors worked in Bayernoil's favor for its 53,000 barrels/day ULSD hydrotreater revamp project, including available, competitive production capacity at a nearby reactor fabricator, and relatively low-cost waterborne transport to a port near the refinery site.

Relatively large savings also accrued because Bayernoil handled its own engineering, procurement and construction (EPC) rather than farming this out to an outside company. This alone can dramatically cut the cost of a reactor project, H2-Advance's Chou told us.

The 450-ton, 18-feet diameter, 70-feet tall reactor vessel cost $2.06 million--a relative bargain, according to Chou and independent refinery EPC experts with whom we consulted. Clever internals allowed cutting reactor height even while increasing the catalyst load. The design should preserve Bayernoil's expected catalyst cycle length--even including the 36% regenerated catalyst loading, Chou said.

Following reactor design completion in December 2000, reactor start-up began in May 2002, in advance of favorable German tax credits for 10-ppm ULSD.

Chou told us that in his prior 21-year career as a plant design and reactor internals specialist at Mobil, his experience indicated that such a reactor vessel normally would cost about $3.5 million, plus between $1.2 million-$l.5 million for internals.

However, Bayernoil was able to use an upgraded Cr-Mo alloy material that reduced vessel weight and cost. Bayernoil also benefited from H2-Advance's bid-quality reactor design package, which doesn't provide a lump-sum contractor "any room for price renegotiation," Chou said. "All the mechanical and process specifications were clearly identified-there isn't any flexibility for the vessel fabricator or subcontractor to gain ground during the vessel fabrication or installation phase," he said.

The project also benefited from a favorable currency exchange rate (1 Euro/U.S.$0.9 in 2001, versus l Euro/U.S.$l.15 in 2003).

For the Bayernoil project, "the cost for the revamp was much lower than what I initially anticipated," Chou told us in a post-AIChE conference interview. "However, the low project cost was contributed by the good timing and competitive bidding process.

"Installed cost for the new reactor including internals is about $3.6 million including about $900,000 for reactor transportation and $640,000 for reactor internals."

"However, these costs exclude reactor foundation, erection, insulation and instrumentation," Chou said.

As this revamp project was relatively "clean-cut," this meant "minimal bulk, indirect engineering cost, so percent contingency assigned to the project can be near-zero," he said.

The project also counted upon H2-Advance's optimized trickle-bed design to improve hydrodesulfurization performance, Rost and Chou explain in their AIChE paper. "This state-of-the-art reactor design demonstrates that it is entirely possible to design a trickle-bed reactor with an inside diameter of 18-feet or larger that still provides minimal channeling for a ULSD project," they said. Such projects require careful quench-zone designs, where big problems can arise.

"Many conventional quench zone designs require excessive reactor height, resulting in high capital investment. H2-Advance's quench-zone design reduces the capital requirement of the reactor up to 10%," by utilizing a "novel quench nozzle and mixing chamber design to reduce the quench zone height while improving thermal equilibrium achievable between the quench gas and the higher temperature fluids coming from the bed above," they explained.

What's more, a "good distributor tray with downcomer design can provide liquid distribution uniformity sufficient for ULSD in a world-scale trickle-bed reactor even if tray levelness is modestly off-spec," the partners found.

The final $106/bpsd revamp cost includes the new reactor, two new shells for feed/effluent exchangers and piping modifications for recycle and make-up gas compressors, but doesn't include costs for process unit reliability improvement items nor the unit-throughput increase from 45,000 to 53,000 barrels/day.