regas LNG- add inerts using supercharged boiler?

[davefitz]

In most cases, countries that import LNG must add inerts ( N2 + CO2) during re-gas operations to lower the Wobbe Index closer to historic domestic natural gas levels. The current method of generating N2 is either cryogenic distillation or PSA- both are very expensive propositions.

It is hypothesized that one can generate a stream of high pressure inerts , approximate concentrations of 1% O2 ,20% CO2, and 79% N2 if one (a) compresses air to 18 bar (b) comusts the air in a supercharged boiler, burning natural gas with a 1% O2 exhaust gas setpoint (c) cools gas with supercharged boiler and aux water heater and (d) removes water via dessicants. The gnerated steam or hot water can be used to aid in heting the LNG to re-gas conditions or to drive a power turbine.

Kindly advise if this has been accomplished already, or been disproven as competitive.

 

[jbmr]

Dave,

This is a very interesting option and I had never heard of it before, but I have to admit that my LNG experience is limited.

Just a question: In the "classical" inert injection process, nitrogen is injected in the recondenser i.e. the low pressure / cryogenic part of the regassification plant. In this case I don't think this is possible because the CO2 would freeze so I guess you inject the inert at the outlet of the vapourizers which is at high pressure and ambient temperature.
Can you generate flue gas directly at the natural gas export pressure (typically 1000 to 1500psi) or do you need additional flue gas compressors?

Thank you

 

[davefitz]

jbmr,

I just realized that with the availability of local cryogenic fluid , it does not make a lot of sense to combust the fuel gas at elevated pressure in a supercharged boiler.

The flue gas can be generated in a normal atmosphehic boiler,and the exhaust gases further cooled to low temps with a HX exhanging heat with the cryogenic fluids. The cooled gases can then be passed thru dessicants and compressed, and the use of cooling prior to compression will reduce the compressor power requirement ( dried air can be cooled to very low temps prior to compression , as the fear of icing is lost). This will be a lower power input than the case of compressing ambient air to higher pressures. Further treatment at 1500 psi pressure can be done to redcuce sizing of dessicants or O2 scavengers.

 

[dcasto]

Why not just inject air. In Colorado the utility injects air to 10%. There is no problem with explosive limits, the air is dried so there is no corrosion.

 

[davefitz]

feedstock users cannot have air in the fuel gas- it affects catalysts or product quality.

 

[davefitz]

Back of the envelope calcs show that a regas terminal would need to burn 0.5% of product fuel gas in the boiler to generate 4% inert ( N2 + Ar + CO2)content of the regas LNG.

Steam generated in the boiler would be used to drive a steam turbine to provide power to drive the compressors for the inerts supply to 1500 psi.

Boiler exhaust at 220 F would be cooled to 60 F using a (condensing) HX that exchanges heat with the LNG that is being regasified. The 19% molar H2O boiler exhaust would be dried to 1.74% H2O in this HX. Further drying using liquid dessicants can be accomplished prior to cooling the exhaust to ( - 200 F) prior to compression. See EEC and DOE projects related to use of liquid dessicants for IGCC and clean coal projects for liquid dessicant performance.

Compression to 1500 psi with 2 stages of intercooling can reduce the comrpessor HP requirement- the ex gases can be cooled to very low temps in these intercoolers if the gases are dry and we make use of available local cyrogenic fluids. Compressors must cryo type compressors.

 

[davefitz]

To minimize O2 content of the boiler's exhaust gases, the boiler would NOT be balanced draft ( ie, no ID fan), and the air heater (if one is used) would be zero leakage type ( ie, heat pipe or tubular air heater). An O2 content of 0.5% molar at the boiler exhaust would be feasible.

If the fuel gas is regassed LNG without odorant, then its zero sulphur content would allow reduced final exhaust gas temperatures without corrosion concerns at the condensing econoizer or air heater outlet .

 

[rmw]

What are you doing about NOX formation during combustion in this supercharged boiler as pertains to NOX limitations in the flue gas all the while?

rmw

 

[davefitz]

It is not neccesary to use the supercharged boiler if you have cryogenic LNG at -260 F locally available, which needs to be heated in the regas process anyway. You can use a conventional atmospheric boiler, then cool the exhaust gases to -200 F ( via an indirect heat exchanger with the LNG) prior to compression of exhaust gases to 1500 psi . This would lower the compressor power requirement significantly,as well as remove H20 as condensate .

Nox can be reduced in the boiler as with normal atmospheric boilers- Lo-Nox burners, flue gas recirculation, possibly SCR but not sure if NH3 slip affects downstream processes.

 

[I2P]

Friends,
Similar to the issue of injecting N2 in LNG, is one of maintaining Wobbe Index of natural gas. I am working in a natural gas processing facility and currently we are having a higher Wobbe Index tyhan allowed by the specs. I have read of injecting Nitrogen in LNG regassing terminals but I would like to know whether this is practised in natural gas processing plants or not? We do not produce LNG but only compressed natural gas.
If at all one does inject nitrogen, how does it affect the calorifc value? How to best estimate the requirement of amount of Nitrogen which needs to be injected per metre cube of natural gas?
I have read in a few reports that increasing the N2 by 1% decreaes the Wobbe index by 1.6%. How do we calculate this.
Our facility is seriously considering injection of inert since we stand to gain revenue due to offspec gas. Also since there is no market for extracted NGLs or LPG that route is also not economically viable.
Thanks to all in advance.

 

[davefitz]

I2P:

You would calculate the impact of N2 injection on Wobbe Index simply by using teh definition of Wobbe Index. One definition is WI=HHV/SQRT(SG) at STP. SG is specific gravity relative to air. Adding N2 to 1% molar leads to a 1% drop in HHV, while also increasing the SG , so the compound affect could well be 1.6% drop in WI.

One problem we would like to address is to minimze the cost of the inerts- generating N2 via cryogenic or PSA is costly both in initial capital and energy consumption. Simply burning some product gas in a conventional boiler and then cooling the exhaust gas to -100F and also removing H2O seems to be far less costly and more energy efficient.

To completely condition the LNG to match doemstic gas quality , regardless of LNG source, one would use both a cold gas extraction of NGL's ( such as the AMEC Paragon process AlphaSimplex) and some degree of inert addition.

A check of commodity prices for LPG and NGL's indicates they are currently priced at a 15% premium over Henry Hub gas prices, and if the regas terminal is local to a port( which by definition they all regas terminals associated with imported LNG are) then the NGL's can be marketed in the global NGL market. According to this month's Oil and Gas Journal article discussing the AlphaSimplex process, rate of returns of 30-45% are calculated for such additions to the regas terminal.

At this point, it sems both technically and economically optimal to provide the regas terminal with both the cold extraction and inert injection, and teh root cause for why the regas developers refuse to consider this remains a mystery.