Knowledge Base
1. When in doubt, ask Jesus what he would do.
Training edit
- Sample Systems
- Present typical system photos
- "Jargon" "lingo"
- Basic GC and stream composition for natural gas
Gas compositions edit
- Phase diagram
- There is -2 degrees to -10 degrees of disagreement between GERG and SRK @ 270 psi and 70 F
- HCDP problem
- orifice plate has liquids on it
- compressor freezes at above-freezing temperatures
- liquid comes out everywhere
- the HCDP is higher than believed
- Gas composition must be known to C9+
- Traditionally all that was available was C6+
- The uncertainty in the stream comp values is at least the same as the calibration gases for the GC
- +/- 2%?
- Would increase to 4% in the heavies, perhaps
- Only 1 decimal place can be considered, max
- Compositions are not constant
- Overall dewpoint fluctuates by +/- 20 degrees F
- Well aging changes composition
- The GC takes a pressure cut before the analysis -> many errors before
- Spot samples are snapshots - The trend is from 1 analysis per month to 1 analysis per year over the past 30 years
- If C6+ is available, adding 30 degrees to the dewpoint is considered conservative (I assume this is when using C6+ = C6)
- If C6+ < 0.02%, this is not required
- Questions
- What is the correlation between the HCD and Wobbe/HHV/Btu etc.
- HCDP dewpoint measurement technologies
- Automated Chilled Mirror
- Laser Scattering
- Michell
- Ametek
- CEIRS
- ZEGAZ
- Laser Scattering
- Automated Chilled Mirror
- Moisture dewpoint - concentration correlation
- ASTM D5503
- What correlation does FlashCalc use?
- ASTM D1142 (1)
- ASTM D1142/IGT-8/Bukacek (2)
- ISO 18453
- Vapor pressure equations vs NG correlation
Materials edit
- 316 SS has carbon content of 0.08% max
- 316L SS has carbon content of 0.03% max
Stuff edit
- Pressure regulation is a throttling process
- Joule-Thompson effect in natural gas is 6-7.5 degrees F for every 100 psi of pressure drop
- Gas prices around the world
- US $3
- Spain $11
- China $15
- Latin America $14
- Wobbe Index (Wow-bee)
- Shale Plays
- Marcellus = traditional gas
- Utica = rich
- Hainesville = traditional with H2S
- Eagle Ford, Bakken = oil-producing, very rich.
- LNG Sampling Techniques
- NBS: LNG Measurement 1985
- ISO 10715-2001
- BS EN ISO 12838-2001
- API 14.1-2006
- Stable Sampling Time
- Pricing for CSA field certification is $7,000 per visit
- Pricing for Cell+Laser Board: ETN Pricing x 1.6
- The %-level H2S application makes use of a backpressure regulator.
- Heated Sample Lines are always recommended
- For every 50 degrees of temperature increase (F), we double the rate of desorption and halve the rate of adsorption.
- H2S experiences chemical adsorption in SS
- Doubling pressure doubles density, doubles adsorption and halves velocity at constant flow rate.
- Gas cylinder standards
- Mixing tolerance
- Analytical tolerance
Sample Systems edit
- Standards
- GPA 2166
- API 14.1
- ISO 10715
- ASTM D-5287
- Design considerations
- Component volume
- probe location
- sample pressures
- tubing length
- tubing diameter
- type and location of sample
- Start with the end in mind
- What is the primary purpose of the sample?
- Accuracy - representative - repeatability
- Utilities-terrain-climate-class location
- lot size-all season access
- Who is impacted by poor sampling?
- Project manager -> Install Cost, cost of bad analysis
- Operations Manager -> ongoing maintenance
- Operating Technician -> Time, frustration
- Manufacturer/rep -> Quoting, service
- Tubing
- For a 3.5 foot probe, 1/2" sch 40 tubing, 3.5 seconds are required.
- For filtration, 1.5" x 7", 6.0 seconds are required.
- For the sample tubing, 150', 1/4" sch 40 tubing, 22 seconds are required.
- Heat trace cycle time can cause H2O readings to fluctuate by up to 2.5 lbs in amplitude; electro-polishing fixes this.
- Use short sample lines
- Reduce high-pressure regions to reduce adsorption
- maximize sample flow to reach equilibrium more quickly
- Tubing treatments in decreasing propensity to adsorb/desorb:
- Conventional 316SS
- Add Silcosteel
- Add chemical passivation
- Add electropolishing
- Add eletropolishing and sulfinert
Corrections edit
- ReSS2100 manual
- every page says H2S analzyer.
- pp.3-27 Val X Concentration
- "When procuring a gas standard, make sure the background gas is
that specified or a mix that closely resembles the contents of the process stream and have the gas standard certified to better than the specified precision of the analyzer, if possible."
- But, ARM9s autovalidate in a nitrogen background in most cases.
- And table 2 analyzers validate in a nitrogen background.
Documentation edit
History edit
El Paso Pipeline Disaster edit
http://www.ntsb.gov/doclib/reports/2003/PAR0301.pdf
http://www.justice.gov/enrd/ENRD_Assets/carlsbad.jpg http://www.constructionequipmentguide.com/scripts/asp_resize_edit.asp?id=5639&width=400 http://www.ndtcabin.com/articles/images/new_mexico_pipeliine.jpg
Corrosion on the internal wall of a natural gas pipeline can occur when the pipe wall is exposed to water and contaminants in the gas, such as O2, H2S, CO2, or chlorides. The nature and extent of the corrosion damage that may occur are functions of the concentration and particular combinations of these various corrosive constituents within the pipe, as well as of the operating conditions of the pipeline. For example, gas velocity and temperature in the pipeline play a significant role in determining if and where corrosion damage may occur. In other words, a particular gas composition may cause corrosion under some operating conditions but not others. Therefore, it would be difficult to develop a precise definition of the term ìcorrosive gasî that would be universally applicable under all operating conditions.
Pecos River Compression Station 32.06373, -104.01856
Pipeline service bridge 32.063642, -104.005562
Block Valves 32.062437, -104.000979
Eunice Plant 32.5142, -103.2857
Keystone Compressor Station 31.94977, -103.10870
Validation edit
- Is a ho
- fo sho
- With standard (calibration?) matrix vs with a second validation matrix
- in pure methane vs in pure nitrogen vs is anything else possible?
- Manual validation
- via external port with internal ball valve vs. 3-way valve in the sample line
- Auto validation
- via internal permeation device
- via external port with internal switching valves
- solenoid valves vs air-operated valves actuated by solenoid valves
- Validation gases required for any field visits, repairs, commissioning. validation gases are specified in technical section.
Applications edit
Natural Gas edit
Gas Gathering edit
- There is gas processing that goes on at these locations, but they aren't strictly considered a Gas Processing location, which is usually a large plant that deals with fractionation of rich natural gas into light hydrocarbon liquids, and residue gas for sales or further processing.
- The measurement should generally be checking the efficiency of the dehydrator, rather than measuring wet raw gas coming from a separator.
- The streams will usually be water-saturated on their way into the plant where moisture will then be removed using Glycol or Mol Sieve dehydrators (the latter are more common at the bigger gas processing facilities).
Order Entry edit
Clean order checklist for web edit
- Inside Sales Contact
- Supporting Documents (attach or list)
- Quote number
- SpectraSensors or sales channel quote number
- PO date
- Bill-to address
- Payment terms
- Commercial contact
- Shipping terms
- Ship to address
- Ship Via
- Shipping contact
- Salesperson (submitting)
- Partial shipments allowed?
- End customer
- End destination
- Discounting (describe)
- Comments
Quoting edit
RFQ checklist edit
- Receive RFQ
- Does it match the app note?
- Stream Composition
- Range
- Repeatability
- Detection_limit#Instrument_detection_limit (or MDL, IDL, LDL, etc)
- Response Time
Spare Parts edit
Spare Parts for SS2100, SS2100a, SS2100i-1, and SS2100i-2 (Powerpoint)
- Provide no spare parts lists unless they are released (i.e. part of the collateral for a released product), as attachments
- Provide a "consumables kit" specific to that custom analyzer, at 8.5% of list price of analyzer system (analyzer electronics + SCS), which includes the wear and tear spares needed for 2 years of operation (or alternatively, for the warranty period), including:
- Filter elements
- O-rings
- Electrical components having failure rates higher than once per 2 years
Stream Compositions edit
- Why does my stream add up to less than 100%?
A component is missing.
Someone messed up when entering the component concentrations.
Someone messed up when averaging or otherwise crunching the values.
Only the dry basis is shown (and not normalized). The missing quantity could represent the water content.
- Significant Figures
- Variability
- C2=? C2*?
Reps edit
Discounts & Commissions edit
Factory Repairs edit
- On some rep contracts a 10% commission is to be paid out on factory repairs (that result from their activity?). Not present on all contracts at first glance. This would apply to any re-calibrations.
H2S edit
Scrubber edit
Limitations edit
- As to moisture, for differential system, to protect the scrubber, I usually recommend limiting H2O below 0.5%. stretch to 2% only if really necessary. For no differential system, better to limit H2O below a couple of % to minimize the collisional broadening effect of H2O on H2S measurement.
-SL
Response time edit
From SM, 1/feb/14: In many cases we really don't have a good understanding of what our response time. On the H2S in natural gas analyzers we did a lot of work on it which resulted in the development of RCM, as well as some other improvements which decrease the scrub time. The product presentations outline the results.
Where we really didn't have an idea what the response time was but we wanted to provide a data sheet, we have fallen back on this parameter called "update time" which is simply the rate that the spectrometer is making measurements. Typically that's going to be 1 second for non-differential and 16 seconds for differential.
The T90 response rate as you described is another thing as you described. The T90 (or time to 90 percent of change) is a common way to measure response time. The reason T90 is used is explained here. http://www.cambustion.com/sites/default/files/instruments/DMS500/timeresponsevsdatarate.pdf
The T90 will be faster if the step size is larger. Also, it depends on the flow rate and the length of the sample tube and volume of the sample system and the cell. All of those things can be fixed for a specific application except maybe the step size.
Most importantly when it comes to differential systems, the T90 will depend on whether there's a dry cycle included - ie it will be a lot longer. We can debate about whether the T90 should include the dry cycle but in my opinion, it should because that's the worst case.
As for trace H2S response time I can tell you how fast we can respond to specific step changes and concentrations where it has been tested. In certain cases we can achieve a 10 second response time even with a dry cycle but it’s probably a T50. The reason we have used T50 in natural gas is because in many cases the customer is not concerned with T90, they are concerned with how long it takes to hit 4 ppmv. For example, if you have a step change from 3.5 ppmv to 4.5 ppmv, then 4 ppmv is at T50. However achieving a 10 second T90 response time with a dry cycle would be very unlikely. We would need to pick a large step change and then test it.
In the end T90 is a complex performance specification that requires thought and decision about what to claim, so we have not yet introduced it into the product data sheets. I am not arguing that it shouldn’t be in the PDS, just that it requires some thought and decision and as you say – consistency across all of the products we sell.