Duty Standby Regime

In the world of process plant, redundant system is being installed. Equipment such as PLC or Control System runs on a full time online fall back redundancy system. Whereas most mechanical equipment runs on a variety of duty-standby arrangement.

In my view, the most effective duty-standby arrangement has to be judged by the engineer. The question to ask is, what is the dominant failure mode. If it is a random failure mode that is dominating, there is very low risk of increase of failure for switching them every fixed period of time. If the dominant failure mode is in the realm of wear, then one might want to consider off-setting the operational hours so they don't have a perfect storm resulting in plant downtime. If the dominant failure mode is false brinelling, one might want to consider running all the redundant equipment at partial load instead!

Managing the depth of RCM

RCM stands for Reliability Centred Maintenance. It is a process whereby a series of questions are structurally raised to conclude a maintenance requirement. It can be as tedious as a full blown RCM where you are moving on average about 3-4 Failure Modes an hour in a RCM workshop to a quick straight forward 30-50 Failure Modes an hour in a peer review workshop.

How deep to go for the RCM analysis? There're a few things to consider when making this call.

1. How skilled are your maintenance team in addressing the Failure Modes? There is no point going into too detail if your trades does not share the understanding and knowledge. For example, carrying out vibration analysis on an equipment without a skilled person is useless. No one will be able to interpret the data and put it to good use. Your strategy would then have to be fine tuned to fixed-time replacement on an optimized shutdown interval.
   
2. How critical is the equipment? The more critical it is, the more time should be invested towards making it performing reliably.
   
3. What is the current state of the maintenance strategy? Is it running reliably? If it is, are we seeing potential Failure Modes that we are not addressing? A peer review to close the gap in the strategy is sufficient in this case. If the equipment is not reliable to start with, it may require a full blown RCM from scratch.
   
4. There will be times where you run into a highly critical equipment but yet the Failure Modes are highly unlikely. The facilitator or reliability engineer would have to make the call whether a full blown RCM is worthwhile or manage the risk with a peer review process to ensure all gaps in strategies are covered. This require local plant experience that none of your external consultants have. Re-emphasize, invest in your reliability team!
   
5. ???

A note to reliability managers out there, if a person pitch you they can deliver RCM workshop at 15 Failure Modes an hour, be very wary about it. You get what you pay for. Quality takes time and it is inevitable in RCM! Again, your best value is through having invest in a very good reliability engineer on your side. After all, the RCM databases will still require maintenance and update in-house, unless you are ready to pay the continual work from the consultancy.

Post-Operational Readiness Project (Continuous Improvement)

With the growing world, the mega projects keep coming. With mega projects comes the need for certainty to the investor. In the world of plant maintenance, the certainty comes from Operational Readiness Project. To some, it is a common phase of a project, to some parts of the world, it is an alien that they have never heard before, that includes experienced multinational EPCM contractors.

If you are deciding on the OPEX of a project, and your EPC or EPCM contractor gives you a cost projection that says, oh, it's 3-5% of capital cost. Ask them where that figure comes from, chances are they'll say it's an estimates from historical data. I can also tell you the figure is INCORRECT. Why? It is because your operating cost depends on the quality of your equipment too! Not in linear, or able to be defined by any mathematical algorithm model! It has to be painstakingly compiled, equipment by equipment, building up to the complete plant! A simplified example, a Japanese car OPEX will be very different from a German car OPEX. They have different service intervals, differing coping ability in operating bandwidth & context, difference in material cost, different in complexity. In some cases, higher capital upfront is justifiable! This has to be evaluated on a case by case basis in details.

Planning and budgeting an Operational Readiness to get your master data up to standard, and having all your equipment registered is essential to every process plant. Once up and running, the OEM recommended maintenance plan has to be put in. No, the process does not end here. You need your reliability team to continually manage the plant changes, and it has to update the master data to reflect the changes occur over-time. As the equipment fails in service, reliability engineers need to assess and evaluate your maintenance task to optimize the time and cost for improvement in reliability, increase in production and reduction in cost. This is an on-going task that cannot be neglected.

From experience, a neglected Master Data can cost upwards of $10 million dollars or more to clean up in a brownfield project and a trailing $1 million dollars a year of labour to keep it clean as they go. If a dedicated person acting as the gate-keeper since day 1 at a price of $150,000 a year, imagine how much less money needs to be re-invested to maintain the original business case projected reliability figures!

A lesson for the Executives out there, Operational Readiness does not guarantee you the reliability outcome. Reliability is a culture and an on-going continuous process. Invest in your reliability team!

Reliability Modelling

I have been working full on until New Year building component library for a consultancy and running reliability modelling. I have run into a lot of issues with the model and would like to share them to promote an understanding.

The very first thing you do as an engineer is always question the validity of those data you acquire. In summary, I would not recommend doing reliability modelling. My personal opinion is, it is a waste of time, effort and money. If you are looking at doing reliability modelling, chances are, your existing plant reliability is not great and your reliability knowledge is not comprehensive. For the accuracy you get, you are better off with a 0.9 factor of industry average reliability figure. No modelling out there I have seen is accurate enough for any good use. If you are doubtful on the quality of people you are able to hire into the maintenance team, use a factor of 0.7 and you should have a somewhat conservative availability figure. Yes it looks ugly, yes it looks unrealistically low, but I'm sorry to say, that is reality of the availability and reliability figure you should expect for saving cost hiring cheap people. I could not emphasize enough, good asset management and reliability starts with good people.

Back onto the topic I was suppose to be writing about - Limitations of reliability model. Firstly, ALL reliability model I have seen is designed in series. It is all well and good if your process is in series like a simple production line of a simple mine site, if you have a complicated processing plant, your reliability model will not do. In fact, there's so much work trying to design the model to fit your plant, it is just not worth the effort. Unless there's a free template already setup similar to your plant and takes just a little bit of effort to patch up, there's no point going down this path.

Secondly, in a complex process plant, you will have varying equipment MTBF and MTTR. Every plant's figure is unique. For the model to be accurate enough to be of any use, it has to be from your plant, your production forecast, you historical availability, and reliability. This is because as the errors build up in the reliability model, the final result is again of not much use to you as the owner.

Thirdly, a complex equipment in a complex plant will have a long list of failure modes to prevent. Some of these failure modes will be attended to in one work task and reset their likelihood of occurrence and budgeted life. None of the reliability model I seen cater for this.

With this three fundamental issues in modelling unresolved, I would not recommend any company looking at carrying out the modelling without understanding the limitations of it.

Plant Mothballing

Read an article in Reliable Plant newsletter today. Made me aware of some work involved prior to mothballing a plant. Good start to that area of practice anyway. :)

http://www.reliableplant.com/Read/28796/do-before-plant-closure

What You Should Do Before a Plant Closure

Bernie Price

It’s happened … the announcement that a major portion of your facility is being closed for the foreseeable future. What do you do next?

1. Accept the news that your plant is facing an impending plant shutdown. It is not necessarily a “knockout” for the plant or your career. Remember, in the often uneven battleground called the global marketplace, just about anything can happen. Be ready to get up and start fighting again.
2. Designate responsibility to an individual for writing a list of possible scenarios. The individual should have enough clout to implement the chosen strategy, if necessary.
3. Go to the top of the company and request that sufficient funds be made available to execute the initial shutdown and preservation strategy.
4. Choose the right type of long-term equipment caretakers. Those selected are often security or ex-supervisory types rather than experienced operator/craftsmen with intimate knowledge of the equipment.
5. Don’t allow critical components to be pirated (stolen for use elsewhere) if part of a larger plant.
6. Remove all process materials. Even innocuous materials left in the unit in the long term will likely cost five times more than at the initial shutdown. The current operations people are familiar with all the hazards.
7. Seek expert advice on equipment preservation resulting in not getting the best bang for the buck.
8. Involve the hourly workforce in the shutdown and mothball plan. Almost unbelievably, our recent experience has been that if the decision to shut down at some future date has been made, then involving operators and mechanics can very much improve both the quality of the shutdown plan and its execution.
9. Not only record but clearly and physically mark what has been done to preserve the item of equipment during deactivation. The reactivating crew (probably a different group of people) can easily miss that a filter, line blind, internal component, etc., has been removed or added with serious consequences at a future start-up.

In our experience, idle plants with small crews operating at a very relaxed tempo can be dangerous places. Make sure safety programs and routine audits are kept active to avoid accidents.
Just as with any critical situation, a long-term strategic approach coupled with a series of medium-term tactics and detailed plans are needed. You should also consider how long the shutdown is probably going to last (guesstimate) and whether the plant will most likely be restarted, sold as a complete unit or sold piecemeal.
Examine every item or class of equipment individually and write a specific initial storage/mothball technique plus a methodology for ongoing maintenance.
For the purposes of this article, let’s consider an item of equipment or a whole plant that might restart as early as six to nine months but could also be several years.
Unused plants and equipment that are properly prepared for shutdown and left in fairly good condition can retain much of their value. However, if a plant is left “as is” and allowed to deteriorate, as is normally the case, much of it can be fit only as scrap in just a matter of months. Engaging in a well-planned process of deactivation/mothballing can be worthwhile either way, whether it should ever be reactivated or just sold for its second-hand value at some future point.

Materials and Equipment You Will Need

Having a clear view of how the constant foes of galvanic/bio corrosion, mold, mildew, etc., can be mitigated if not defeated is essential. Although much will depend on local conditions, the wetter and colder situations are much more challenging in terms of handling humidity, while blowing dust is an issue for those in the high desert regions. For this article, we will consider a central United States or European location.
A useful analogy in developing a strategy is to compare what it takes to maintain fire. In the case of fire, there are three essential legs: heat, a fuel source and oxygen. Likewise, age-related deterioration involves a driving force (such as galvanic action), a conducting medium (electrolyte) and oxygen. The fundamental approach to stopping or slowing this age-related deterioration is to remove one or more of the three legs.
In simple terms, we aim to do the following:

• Separate dissimilar metals.
• Protect surfaces that could be attacked, even with a covering of only a few molecules thick.
• Dry out or remove the conducting medium (electrolyte — air or gas). Corrosion cannot occur when parts are stored in environments where the relative humidity is held below 40 percent.
• Remove any oxygen or sources of chemical or biological attack.

The materials we can use are:

• Liquid protective waxes and liquid polyvinyl chloride (PVC) coating — These can be sprayed on any clean, dry surface to protect them. Wherever it is applied, PVC will form a tough, flexible, waterproof skin that will withstand the extremes of temperature, thermal shock, differential substrate movement and impingement even when sprayed on webbing to form a cocoon.
• Volatile corrosion inhibitors (VCIs) — These generate protective vapors even at room temperatures. They come in a number of convenient forms, including time-release vaporizers, sprays, plastic bags and films, powders, oil additives and coatings. They are adsorbed onto the metallic surfaces of the equipment (just a few molecules thick), where they can prevent corrosion for up to two years. While most VCIs are environmentally friendly and create no safety hazards for employees, there are some that are suspected of being harmful. Most contain no toxic substances, such as nitrates or chromates. (Note: Volatile organic compounds should not be used in combination with a desiccant.)
• Vapor space inhibitor (VSI) — This is an oily concentrate that can be added to lubricating oil systems (internal combustion engines, etc.) when equipment is not going to be completely filled.
• Heat-shrinkable plastic films — These are ideal for enclosing individual machines that have been cleaned and dried and have internal desiccants added.
• VCI-covered polythene films — These are used to wrap individual smaller components.
• Chemical oxygen scavengers — These are frequently added to fresh water used to displace more corrosive liquid in systems that can’t be effectively cleaned or dried out.
• Chemical inhibitors — These are added to liquids and chemicals and are designed to remove unwanted products while preferentially inhibiting their attack on the body of the container. (Antifreeze sometimes used in this process contains them.)
• Desiccants — These include numerous substances (solids) that absorb water from gases (air) or liquids.
• Biocides — These are used to prevent microbial growths in water and fuels such as gasoline and diesel fuel.
• Light waxes — These are used as surface protectors for metals.
• Sacrificial Anodes — These are used in tanks that cannot be drained of their contents.

The primary pieces of equipment are dehumidifiers. These are available in two forms: those that work on the refrigeration principle and those that use two-cycle rotary (wheel) heated desiccant absorption.

Strategies by Equipment Class

Before considering individual techniques, make a best guess of the duration and whether it is going to be an “attended monitored” lay-up or a “walk-away” lay-up. This article is simply a guide and is not intended to be totally comprehensive and detailed.

Tanks, Pressure Vessels and Pipework

It is essential that tanks, pressure vessels and pipework be left as clean and dry as possible. Insert line blinds to create manageable zones that can be slightly pressurized (0.5 psig+) using nitrogen or dry air. Include some small flow and arrange for some simple telltale mechanism to show pressure flow and the level of humidity (indicator cards). For large enclosures, use a commercial dehumidifier of an appropriate capacity. For vessels, tanks and containments that must be kept full of liquid, some form of oxygen scavenger or anti-biological growth chemical can be used (see boilers). If a pipework system contains any traps, have its internals removed and clear all strainers.

Boilers

Boilers can be laid up using either the long-term dry method or the hydrazine wet lay-up method, which involves leaving the wet side (boiler, economizer and super heater) full of feed-treated water. The feed water is dosed with 15 percent hydrazine and then pH-adjusted to raise the alkalinity to a minimum pH of 8.3. The fire side is supplied with heated air, with desiccant as a backup. Both water-side and fire-side points should have new gaskets, except for furnace hot-air entry inspection and exit points.

Pumps, Engines, Compressors and Machinery

To minimize internal corrosion, close off all vents and openings, and completely fill the casing with the manufacturer’s recommended lubricant. Alternatively, add a volatile corrosion inhibitor in the correct proportion to the lubricating oil. For large compressors, turbines, etc., first centrifuge/circulate the existing oil using a portable filtration cart through water-absorbing filter elements to remove any free water. For diesel and gasoline engines, drain the fuel systems and add biocide to the remaining fuel. To prevent external corrosion, if unpainted, one of the recommended spray-on coatings should be used (either a light wax or liquid PVC).

Instruments/Controls

Maintaining the driest possible conditions for both electronics and external field devices, including sensors, transmitters and valves, can be achieved by strategic placement of desiccant packages and sealing the enclosures. This should be supplemented by placing small containers of VCI powder wherever possible. These will not adversely affect electronics. Instruments that normally would be in contact with the process materials should be removed, cleaned, protected and marked for immediate local storage.

Electrical Enclosures

Seal and insert bags or wraps of desiccants and containers of volatile corrosion inhibitors. Alternatively, heat using individual strip or built-in heaters.

Motors and Generators

Clean the exterior, grease and apply a protective covering. If completely sealed, add packets of desiccant. Lift carbon brushes from commutators/slip rings. Where sleeve-type bearings are fitted, a VSI concentrate should be added to the lubrication system.

Exercising and Monitoring

Depending on the time involved, it will be necessary to periodically exercise equipment by rotating it several times and leaving it at a different (90-degree) angle. Where humidity controls have been set, these need at least weekly monitoring. Where chemical controls are used, these should be checked every three months. Periodic monitoring of motor/generator internal resistance, as well as tank oxygen levels and humidity levels, is necessary for long-term lay-up.

Auxiliaries

In most cases, various fire-protection systems and alarms still need to be maintained and powered up. Fires are common in dried-out cooling towers. If batteries are normally used, disconnect them and smear the terminals with petroleum jelly. Vented-type lead-acid batteries should first be fully charged, then drained and flushed with distilled water.

A Final Note

A recent discussion with two major plant-dismantling/second-hand equipment vendors revealed that currently there are very few people looking for used equipment, and many idle plants are being offered for sale. They reported that when the decision to shut down comes, most companies remove anything that could present an immediate danger but essentially close the doors and walk away from millions of dollars’ worth of equipment. 

Gear Coupling reference 1

Found a useful article in Reliable Plant newsletter today in regards to couplings.

How to Achieve Gear Coupling Reliability

http://www.machinerylubrication.com/Read/28851/gear-coupling-reliability

How to Achieve Gear Coupling Reliability

Randy Riddell, International Paper 

Design, Selection and Sizing

Selecting the correct coupling for the application is critical for gear coupling reliability. Use the following steps to help make the selection process easier:

1. Choose the coupling style and design (Fast’s, Series H or Waldron; flex and rigid halves; close coupled or floating shaft; gear teeth specifications and misalignment requirements).
2. Select the service factor (SF) from the original equipment manufacturer’s (OEM) gear coupling charts. Shock loads or variable loading can cause premature failure if adequate SF is not used. Typical service factors are in the 1.5 to 2.0 range. Some manufacturers may even specify a misalignment factor for gear coupling sizing when higher coupling misalignment is expected.
3. Calculate application torque (T) requirements based on design brake horsepower (BHP), SF and speed.
   
4. Choose a coupling with a torque capacity greater than the torque requirements. Since the service factor is already factored in, there is no reason to add additional capacity.
5. Confirm that the coupling selected has a bore capacity greater than the actual application bore (shaft size). Frequently the maximum bore size will drive the coupling sizing process and even increase the coupling torque capacity two to three times what was previously calculated.
6. Verify the shaft depth available for the coupling hub and compare to the actual hub depth. If the hub is too long, it must be either overhung or machined off. Since the hub to shaft engagement is the same in either method, it is preferred to have the hub machined off due to torsional effects of the overhung hub. If the hub is overhung or cut off, further examination may be necessary to determine if there is enough torque transmission capacity available. The rule of thumb is a 1-to-1 ratio for the hub length to the bore.
7. Check a dynamic balance chart to see if the coupling needs to be balanced. High-speed gear couplings may require balancing.
8. Ensure the coupling will fit around the equipment and guarding. This is typically something that can become an issue when there is a design modification on existing equipment. Guards that allow maintainability will encourage proper maintenance in the long run.

Installation

Some couplings don’t get much of a chance at a decent life due to their installation. Just like other components that experience infant mortality, often times these parts don’t die but are murdered. Certain elements of gear coupling installation must be considered if optimum reliability is to be obtained, including:

• Hub and Sleeve Fits - Determine the type of hub fit (clearance, locational or interference). Higher speed applications should have an adequate interference fit to offset centrifugal force effects on shaft/hub contact pressures. Excessive hub interference fits can lead to hub cracks and hub failure.
• Keys and Keyway Fits - Keyways should have a proper radius to reduce the risk for fatigue cracking. Key lengths should be measured to minimize the coupling imbalance.
• Hub Bore - Ensure the hub bore is concentric to minimize hub runout.
• Hub Installation - Choose proper heating methods so hub material properties are not compromised and select the proper heating magnitude for interference fit hubs so the hub slides easily on the shaft. Never use a hammer to install or remove hubs, as this can cause bearing damage.
• Correct Coupling Gaps - If floating shafts have a small coupling gap, the shafts may impact one another under misalignment as the shaft oscillates during operation.
• Proper Sealing - Always use proper gaskets and O-rings so the lubricant stays in the coupling.
• Alignment - Install the coupling so misalignment stays within manufacturer limits with respect to offset, angular and axial misalignment.
• Fastener Assembly - Choose the correct type of fasteners (fine or coarse, length, exposed, shrouded, etc.) and the proper arrangement. While standard bolts can work, they may put the threads in the shear plane. Coupling bolts need the correct preload, which is accomplished by proper bolt torque methods.
• Lubrication - Get the right product in the right amount at the right time for optimum gear coupling reliability.

Different coupling styles have different lube and bore capacities. (Ref. Kopflex)

Lubrication

Perhaps the most important operating factor for a gear coupling to be reliable is lubrication. Selection of the proper lubricant is the first step. Many coupling manufacturers supply their own lubricants for their couplings. Gear couplings may either be grease- or oil-lubricated depending on the design. Oil-lubricated couplings will not dry out like grease couplings, while Fast-style couplings have smaller bore capacities.
It is fair to say that most gear couplings are grease-lubricated. Coupling greases have special properties, so general-purpose greases should never be used in gear coupling applications. Gear couplings can be subjected to very high centrifugal forces, and oil separation is a critical element of coupling greases. Since greases are comprised of oil and mostly a thickener, special considerations must be made regarding the selection and application of coupling greases.
Soap thickeners typically are heavier than the oils, so centrifugal forces tend to deposit the thickener at the gear teeth. Generally, a grease with a high oil content of high-viscosity oil and a grade 1 rating from the National Lubricating Grease Institute (NLGI) is preferred. A higher consistency grease may be considered for high-speed applications but should be avoided at low-speed applications.
Grease specifications may include speed limits or certain tests such as the K36 separation factor. Any grease will have oil separation based on time, temperature and centrifugal force. The K36 factor determines the maximum oil separation of the grease while running at 36,000 Gs. A K36 factor of 8/24 means the oil separation was 8 percent in 24 hours. In comparison, a grease with a K36 factor of 3/24 would mean that it did not separate as much as the grease with a K36 factor of 8/24.
Higher oil separation is desirable at lower speeds (lower G forces), while lower oil separation is preferred at higher speeds and higher temperatures. High-vibration equipment can also enhance oil separation and induce failures. Studies have even shown that gear coupling wear rates decrease as coupling speeds increase.
The main function of a lubricant in a gear coupling is to reduce the friction between the gear teeth as they slide against each other. The relative motion between the mating gear teeth occurs in the axial direction due to slight shaft misalignment. This motion is oscillatory, low amplitude, relatively high frequency and a function of the magnitude of angular misalignment.
This sliding axial motion between the gear teeth can generate lots of wear if lubrication is not sufficient. This is why the gear coupling lubricant plays such a critical role in the reliability and life of a gear coupling. Poor lubrication between the gear teeth generates higher friction between these teeth, resulting in gear coupling wear, heat generation and high axial loads to mating equipment bearings. The higher axial loads on the bearings will then decrease the life of the equipment.
The pump shown on the left had a dry coupling that was operating in a torque-lock condition and creating high axial forces on the equipment. The coupling was replaced without making any adjustments to the pump or motor. The only change was a coupling with good lubrication, which reduced tooth friction and decreased the axial forces from the coupling to the pump and motor. The result was a noticeable decrease in the operating temperature of the pump bearing.

Maintenance

Maintenance is the final factor to ensure gear coupling reliability for long equipment life. While the first three factors have more to do with a lack of knowledge, maintenance often comes down to a lack of execution. Unfortunately, this requires discipline by operations and maintenance groups as well as managerial courage to dedicate the resources to ensure that it can happen.
Typical recommendations from gear coupling manufacturers require regreasing at a minimum of 12 months. A regreasing procedure would include breaking, cleaning, inspecting and hand-packing the coupling with fresh grease. Using a grease gun typically is not recommended when the coupling has been broken and ready to receive new grease. When a gear coupling is greased through a fitting instead of hand-packing, it can result in overgreasing, and a hydraulic lock condition can occur, causing high axial forces on the equipment. A hydraulic lock condition can even make alignment difficult, as shafts may be hard to turn.

Some applications require regreasing at six months to ensure good reliability. These applications may include high speeds (high G forces), high temperatures, misalignment or vibration. Smaller lube sump capacity can also be a factor in regreasing intervals. However, deciding to go longer than 12 months without grease replenishment on a gear coupling is a high-risk move that is not recommended.
Regular maintenance of gear couplings should involve special care with respect to many of the installation factors discussed previously. When inspecting gaskets and O-rings, ensure the lubricant stays in the coupling until the next maintenance task is scheduled. Grease fittings should be removed before completing maintenance. These fittings have been known to leak lubricant and can hit guarding, causing loss of lubricant. Under high centrifugal forces, the grease must be completely sealed within the coupling. Guarding should also allow enough access so it does not have to be completely removed for normal coupling maintenance.
Remember, reliability is not for the faint of heart. Most all of these factors must be executed correctly to achieve good gear coupling reliability. This is why the work of maintenance and reliability professionals is rarely ever finished.

About the Author

Randy Riddell is a senior mechanical reliability engineer for International Paper in Courtland, Ala. He is a certified lubrication specialist (CLS) by the Society of Tribologists and Lubrication Engineers and a certified level I machinery lubrication technician (MLT) by the International Council for Machinery Lubrication (ICML). He is also a certified maintenance and reliability professional (CMRP) by the Society for Maintenance and Reliability Professionals (SMRP).

Transformer Maintenance - Silicone Oils

This entry is an extension of Transformer Maintenance - Mineral Oils to some extend. Again, this article is not a complete guide. It gives you an overview of how complicated it is. A lot is to be considered in making an informed decision in regards to a transformer especially if it is a highly critical asset.

Silicone oil when new contains a saturated amount of oxygen. In the initial years of operations, carbon monoxide and carbon dioxide will be generated. As the transformer ages, oxygen is depleted, generation of these gasses slows and plateau off after a few years of operations assuming without any faults. The generation rates of these gases should be relatively constant from normal aging after that. Which is why it is very important to start DGA immediately and start plotting the graph curves and track these changes. Without a graph, it is almost impossible to make a judgement.

Comparison of Silicone Oil and Mineral Oil.

1. Silicone oil-filled transformer will have a great deal more CO than normal mineral oil-filled transformers. CO comes from the oil itself and from degradation of paper insulation. It is therefore if DGA indicated little other fault in gas generation besides CO, the only way to tell for certain if CO is coming from paper degradation is through furan analysis. If other gasses are involved, there obviously is a fault and paper degradation was accelerated.
2. Hydrogen level is generally higher comparatively to Mineral oil filled transformer.
3. Due to "fault masking" environment with Silicone oil, DGA lost many of its fault finding capabilities. One exception is acetylene that points to an active arcing. It is then very important to continually track the gas generation rates and operating history. Records, records, records!
4. Oxygen level will be high during new and consumed over its life by the generation of CO and CO2.
5. Any spike in O2, CO2 and N2 after a few plateau reading would very likely indicate a leak to atmosphere.

Due to the rather infancy stage of Silicone Oil usage, these gas limit extracts are use as a reference and will change over time as the world gain more experience dealing with them. Use it with care. This is a Doble 95% Norm limits of 299 operating transformers, which are more conservative in some way than IEEE limits.

Hydrogen                              511ppm
Methane                                134ppm
Ethane                                     26ppm
Ethylene                                  17ppm
Acetylene                                  1ppm
CO                                     1750ppm
CO2                                 15480ppm
Total Combustibles              2000ppm

Keep in mind that the amount of gas is not the key. The key is the generation rate of the gasses. Refer to IEC 60599 for the generation rates. G1 rates should raised concerns along with sampling rates increased and expert opinions seeked. G2 rates should be an immediate extreme concern that the reaching the L3 - high limit of IEEE will happen very quickly. Consideration should be taken to take it offline.

A reference of physical test limits for service-aged silicone fluid
Test                                     Acceptable limits      Unacceptable values indicated         ASTM method
Visual                                  Clear free of particles        Particulates, free water                     D1524, D2129
Dielectric breakdown                     30kV                     Particulates, dissolved water                      D877
Water content max.        70ppm(Doble) 100ppm (IEEE)  Dissolved water contamination              D1533
Power Factor max@25degC         0.2                           Polar/ionic contamination                         D924
Viscosity at 25degC, cSt          47.5-52.5                    Fluid degradation contamination                  D44
Acid number                        0.1(Doble) 0.2(IEEE)  Degradation of cellulose or contamination        D974

Reference of this article:

1. Transformers: Basics, Maintenance, and Diagnostics - Reclamation, US Department of the Interior Bureau of Reclamation, April 2005.
2. Trial-Use Guide for the Interpretation of Gases Generated in Silicone-Immersed Transformers, IEEE P1258, 1999.