Save as pdf document

GMRC Visual Analysis School


Slide 01GMRC-Slide01

Slide 02

Firstly, I need to preface this paper by advising you there is no way that the spark plug can be discussed thoroughly in an hour. So, to supplement this presentation, I am furnishing some of our printed literature that discusses some of the issues important to you in greater detail.

Furthermore, we have an Engineering Manual that you might find educational. Unfortunately, it is in the process of being remodeled, but if you would be interested in seeing it in its old format, please let me know and I shall see to it that you are furnished a copy of the initial edition.

Having said that, here are some of my musings on topics that I thought would be meaningful to you!


Slide 03


The puzzle of the spark plug is not that it does more than one thing during its interval of operation, but that so few engine operators take advantage of the other feature that it offers.

All engine operators take advantage of its principal feature as the indispensable last link in the transfer of electrons from the magneto, through the cylinder head, to the spark gap within the combustion chamber.


Slide 04

And in its most basic sense, that electrical function defines what a spark plug is:

A high electrical voltage “pass-through” providing a spark gap with a controlled rate of thermal transfer.


Slide 05

But the spark plug offers the sophisticated engine operator much more than this well-recognized, basic feature.

For the analytical engine operator, the spark plug functions as a recorder of the combustion events that it experiences during its service life.

And it is a recorder offering such exquisite detail of the combustion environment, to the informed observer, that the major puzzle is the mystery of why so few operators use this analytical feature on a routine basis.

After all, we encounter the proposition that many operators routinely analyze “used” engine lube oils. And yet we find few operators that have any interest in routinely analyzing their “used” spark plugs.

This is the major puzzle, because I think it a truism that lube oil analysis provides no complete picture of the combustion area of the engine.



Slide 06



Slide 07



Slide 08



Slide 09


Slide 10

To the best of my knowledge, the majority of catastrophic wear trends involve the engines in an area from the lower ring land at BDC to the firedeck. There is a large amount of wear evidence that occurs in this area and never reaches the crankcase: it goes out on the exhaust gases. We think it unlikely that conventional lube oil analysis identifies these elements.

And we think it even more unlikely that an oscilloscope will reveal them.

What we think to be of interest is that the spark plug will retain the record of its life in the combustion environment. Pretty accurately, its analysis can identify how it is being used and/or abused. What makes a spark plug such a convenient analytical tool is that it is an engine component that is easily removed from that environment. More importantly, it is an engine component that can be placed within the vacuum chamber of a scanning electron microscope relatively easily, without altering its “as-received” condition.

Because of these realities, for at least twenty-five[25] years, Stitt has been offering such analytical services to the users of its spark plugs. So as to make the analysis meaningful, as well as to render any such submission eligible for our analytical consideration, we have some requirements that the engine operator must fulfill. We think that our requirements reflect common sense and are the only methods by which a complete picture of the engine can be obtained.


Slide 11



Slide 12


A. All of the spark plugs from the engine must be returned, each one of them identified by its engine location.

B. Along with the spark plugs, we require the submission of additional detail[PSAR-FORM].


Slide 13

Upon our receipt of these spark plugs and the PSAR information, these spark plugs enter our analytical process. Here is that process…..

1. Spark plugs are visually examined. Spark plugs exhibiting any “naked eye” discrepancies[more unusual than those illustrated in our VISUAL ANALYSIS manual] are noted for a more microscopic examination.

2. Spark plugs are all measured for gap dimension[s] and examined so as to establish electrode erosion.


Slide 14

Slide 15

3. External seat gaskets are examined, measured for thickness: this provides an insight into cylinder head seating surface conditions as well as installation torque.

Special Note: In virtually 99% of the problem spark plugs that have been/are submitted to us for our analysis, the spark plugs have not been properly installed. Primarily overtorqued, but often untorqued. Either error will result in spark plug problems. In some instances, these installation torque errors can become directly responsible for spark plugs becoming projectiles and advancing the risk to life and limb of operating personnel.[More about this later].

4. Spark plugs are evaluated for electrical behavior using both a VOM and a Megger.

5a. If the visual appearance of the submitted plugs warrants it, then the plugs are color photographed[color-corrected] and more closely examined by an independent metallurgical lab. As a rule, all of the relevant aspects of the plugs are photographed in color. Principally so as to inform the operator as to what he could have seen had he himself looked at these spark plugs carefully.


Slide 16

5b. If any one of the spark plugs exhibit an operating characteristic deserving of a more detailed analysis, those examinations are also conducted by third-parties and a photographic record will be made.


Slide 17

6. For unusual deposits[generally firing-end deposits], those deposits will be subjected to a quantitative element analysis[EDAX] in the vacuum chamber of a scanning electron microscope.


Slide 18

7. Since the element sulfur[S] almost always appears as a deposit element, and since a form of sulfur[eg.,sulfate, sulfanate] is a constituent of most lube oils, if we see any unexplainable erosion of electrodes, we do a qualitative examination of the deposit to see it the sulfur within the deposit is in the sulfide state. If it is, it is strongly suggestive that the spark plug has been subjected to hydrogen sulfide[H2S]-bearing fuel gases. As you might expect, the more active the reaction to the sodium azide, the greater the likelihood that the spark plug has been exposed to unusually high levels of hydrogen sulfide[H2S] in the fuel gas.

GMRC-Slide18Slide 19

8. On occasion, hydrogen sulfide[H2S]-induced damage will so injure spark plug components that a “Naked Eye” exam may conclude that spark plug electrodes were experiencing some form of abnormal thermal distress[melting]. So as to overcome the inadequacies of that level of exam, we subject the supposedly thermally distressed areas to a more molecular level of scrutiny.GMRC-Slide19Slide 20

When this examination reveals the separation of grain boundaries, intragranular attack[corrosion], the conclusion is always that thermal distress from abnormally elevated temperatures has not been the controlling factor in this deterioration.

Because hydrogen sulfide[H2S] can often be a component of the natural gas used as an engine fuel, I would like to pay some special attention to this form of spark plug damage, since it is so often mistaken for thermal distress rather than for what it really is: chemical attack.

Even more to the point, that this form of chemical attack is occurring is very important to determine. An accelerated rate of spark plug electrode erosion or degradation can be considered the “tip of the iceberg” if unexpectedly high quantities of hydrogen sulfide are entering the engine on the fuel gas stream. Just imagine what other, more expensive, more life-threatening components are being driven to failure by their exposure to this very energetic corrosive.


Slide 21

At this time, I would like to share with you our analytical data concerning engine sets of spark plugs where the spark plugs were returned because of assumed thermal distress based on the operator’s “naked eye” observations. What our analytical tools indicated was that hydrogen sulfide[H2S] attack – intragranular corrosion – caused the spark plugs to appear as if thermally distressed.

By the way, as a rule, to the best of our knowledge, none of the instruments that are routinely used to analyze these engines in operation will detect this problem.

GMRC-Slide21Slide 22

That having been said, here is some of the intragranular attack evidence that we have been allowed to see. Based on the “where there is smoke there is fire” understanding, we think the number of spark plugs thought to have been failed for “thermal” reasons is way too large. Recognizing the chemistry of natural gas, especially odorized natural gas, we think that hydrogen sulfide[H2S]-attack may be one of the most unacknowledged of spark plug assassins.

To support that assertion, I submit to you some of the evidence that we have collected. More to the point, I think that you will see that the way we have conducted these examinations reveals avenues for combustion trendline development that mostly goes unexamined by virtually all engine operators.


Slide 23


GMRC-Slide23Slide 24


GMRC-Slide24Slide 25

GMRC-Slide25Slide 26

GMRC-Slide26Slide 27

GMRC-Slide27Slide 28

GMRC-Slide28Slide 29

GMRC-Slide29Slide 30

To conclude this discussion, I want to encourage the practice of scientific method in the operation of these engines.

In many instances, the engine operating personnel focus on a symptom, short spark plug life, rather than the disease[i.e., more organic problems not originating with the spark plug]. The spark plug did not introduce the hydrogen sulfide[H2S] into the combustion environment even though it was adversely affected by that fuel gas constituent.

Almost always, when the spark plug is thought to be “faulty”, the actual “faulting” components of the equation reside elsewhere. At least that is what we have found when we go into the field to assist our customers who report that they are experiencing spark plug problems.

As a matter of policy, when allowed to do so, Stitt personnel conduct “REFERENCE” installations so as to insure that the spark plug is as unaffected by as many problems as we can eliminate before the spark plugs are required to perform. For the purposes of this presentation, we submit that these installation procedures should be the standard, normative set of procedures for all spark plugs in all engines.


GMRC-Slide30Slide 31


GMRC-Slide31Slide 32

GMRC-Slide32Slide 33

GMRC-Slide33Slide 34

GMRC-Slide34Slide 35

GMRC-Slide35Slide 36

GMRC-Slide36Slide 37


As a matter of policy, whenever we receive communication from a customer asking for our recommendations for an engine that is experiencing “spark plug problems”, we try to put personnel at the engine site so as to make a factory-controlled, spark plug installation.

Now, we do this to insure that the spark plugs have the best chance of performing successfully. Most importantly, we do this because our experiences have convinced us that most engines have other deficiencies that compromise spark plug performance: that no spark plug can overcome such deficiencies. At the very minimum, for the spark plug installation to proceed, we analyze the engine so as to determine what deficiencies have to be corrected.

1. If we have the opportunity, we try to analyze the secondary output characteristics of the engine before the engine is brought down and taken out of service. These days, we use Watchdog 1000’s and we store the secondary ignition data for each cylinder[each spark plug].



Slide 38

Any anomalies observed are noted for further examination. Additionally, we attempt to do a visual exam of the secondary side of the ignition so as to ascertain if there are any “in your face” anomalies[i.e., dielectrically punctured, plastic extensions; dielectrically failed secondary leads; arc-tracking, corroded high voltage, coil towers; and arcing over spark plugs].

At this point, I would like to assert that, in many instances, a common sensible, regular, visual examination of the ignition components in operation will reveal the source[s] of operating problems as well as the usage of electronic, diagnostic devices.



Slide 39


Slide 40


Slide 41



Slide 42


Slide 43

2. After the engine is brought down, we disconnect the spark plugs from their upstream components and remove them from the engine. If I am at the site, I like to use a balance-beam style of torque wrench so as to indicate each spark plug’s break-away torque. Many times these break-away torque levels will reveal insights into the prior spark plug installation practices.

GMRC-Slide43Slide 44

3. After the used spark plugs and secondary leads are removed from the engine, we inspect them with the “naked eye”. We might be able to augment the “naked-eye” with a magnifying glass, but a more microscopic exam could not be performed in the field, ordinarily. For that level of investigation, the plugs would have to be returned for a lab exam.

We observe spark plug appearance. We check the used spark plug gaps.
Where those gaps do not correspond to the demand voltages that we measured, we note those secondary/primary components for further evaluation.GMRC-Slide44Slide 45


Slide 46

GMRC-Slide46Slide 47

For example, if we note a secondary circuit where the demand voltage exceeds 20kV, and yet the spark gap dimension measures less than .015″, then we determine that the ignition components upstream of the spark gap require more evaluation so as to determine their adverse influence on this observed, anomalous, high voltage demand.

This means that we shall try and read continuities/resistances for that circuit from the ignition control module, through the primary wiring, then through the coil, then through the secondary lead, and finally through the spark plug.


Slide 48

In most instances of “reported” short spark plug life, upon completing these investigations, we have found that the spark plug problems originate with upstream ignition components[i.e., coils with no continuity, coils with excessive corrosion in the high tension tower, excessive resistance in the primary wiring, no continuity in the secondary leads, insulation failures upstream of the spark plugs, etc].

If any ignition “faults” upstream of a spark plug cannot be corrected[eg, a good coil replacing a bad coil, for instance] then we do not continue with the scheduled spark plug installation.

But, if all the upstream ignition components are OK, then we proceed to the next step.

GMRC-Slide48Slide 49

4. Which is to gage all of the spark plug ports. You need to know that we generally find that the spark plugs’ performance is being compromised because the spark plug ports are out of standard tolerance.GMRC-Slide49Slide 50

4a. An oversize spark plug port, the NO GO goes, will generally be linked to spark plugs that have experienced a faster rate of electrode erosion. This oversize spark plug port condition will inhibit the desired rates of thermal transfer from the spark plug that the engine operating conditions demand.

In the worst case situation, the air gap between the spark plug port threads and the spark plug threads is so great that the whole spark plug blows out of the cylinder head because it is insufficiently thread engaged. Or, the spark plug shell overheats so badly that it reaches a uniform 650 degree F temperature which causes the steel to give up over half its tensile strength, thus allowing for the distinct possibility of the insulator assembly exiting that shell explosively.


Slide 51

4b. An undersize spark plug port, the GO does not go, is what we have found most often, once we started looking at this aspect of the spark plugs’ interface to the engine.

Suffice it to say, we have been forced to conclude that cylinder heads, pre-combustion chambers, all too frequently, are not gage inspected by the companies that fabricate them, that sell them.

Up until a few years ago, I assumed that thread lubricant was applied to spark plugs so as to facilitate their extraction. But after some REFERENCE spark plug installations that I made some time ago, I discovered that thread lubricant was being used to allow the fitting of the spark plug.

In fact, we found a number of engine operating facilities that were not just lubricating the threads so as to effect installation, but were having to fortify the lubricated threads with air-driven, impact wrenches.

These are not good practices. These attempts to overcome a “defective” spark plug PORT thread should destroy critical constructional features of the spark plug.


Slide 52

If a spark plug cannot be fitted into the spark plug port to the seat position by hand, then the spark plug port needs to be evaluated for its compliance with SAE Standards. As a rule, if we cannot accomplish a “by-hand” fitting, then we do not proceed with any factory-sanctioned spark plug installation.
To all the engine operators out there, we would strongly recommend that you assert spark plug port thread specification for the spark plug ports that you purchase. Insist on a certificate that these critical threads have been gaged.

We recommend that no new engine, no remanned engine, no new cylinder head, no remanned cylinder head, no precombustion chamber, no spark plug receiver be accepted by any engine operator until its spark plug port threads have been fairly inspected.


Slide 53

5. If possible, true engine ground is verified. That it exists and that it is provided as specified by the ignition control module manufacturer. Especially if other sophisticated electronic components are part of the engines’ operating scheme.

6. When all the consequential anomalies have been corrected/resolved, then, and only then, do we install the spark plugs and connect them to their “drivers”.

7. Once those connections have been made, and the engine is re-started and loaded, then we re-scope the secondary ignition performance for each secondary circuit.

Though this is too involved a procedure for each spark plug change procedure, at some point this procedure should be documented so as to provide an electro-mechanical benchmark for the subject engine. Candidly, we think that this level of scrutiny should be conducted on all the available operating horsepower on a pipeline at the first opportunity.


Slide 54


Radio Frequency Interference[RFI] aka ElectroMotive Interference[EMI] NULLIFICATION OF ANALYSIS EQUIPMENT.

I think this is a very important issue. That is not well understood.

For some reason, there is a prevailing, conventional wisdom that says that if there is any “hash” on an oscilloscope, that it must be because of the “resistor” status of the spark plug[s] being operated. That is an inaccurate understanding of how RFI/EMI can be attenuated, suppressed by means of any spark plug constructional features.

For any spark plug resistor to have any effect on ignition-generated RFI/EMI the high voltage arc at the spark plug spark gap[s] must be the only high voltage arcing. As a matter of function, using an industrial, low-tension ignition set of components, this should be the only intrinsic arc-site. As a rule, when the arc is struck and sustained across the spark gap, as close to that spark gap as possible, any resistor in the spark plug can only suppress[attenuate] the RFI/EMI developed at that one site.


Slide 55

So, it must be understood that if there is any connection that provokes an arcing connection upstream of any spark gap, then no resistor in a spark plug will have any influence on that RFI/EMI generation site.


Slide 56

GMRC-Slide56Slide 57

In fact, if there are any arc sites upstream of the spark gap, then any spark plug resistance will be completely irrelevant.

Of even greater importance, any high voltage arcing, from any site, should make it virtually impossible to read an oscilloscope.

GMRC-Slide57Slide 58


Slide 59


Twenty-five[25] years ago, Ingersoll-Rand in Painted Post, NY told me that it was a mandatory requirement for spark plugs to be recommended for operation in their KVS, KVR series of engines that they be constructed so as to prevent the spark plug insulators from “blowing-out”, so as to prevent the threat to life and limb less-robustly constructed spark plugs provoked.

GMRC-Slide59Slide 60

In response to that requirement, Champion offered their vision of an insulator-restraining design, the two-piece design[RW83F], BG offered one, and we offered our E72.

Of the three designs, even 25 years later I think it still quite obvious that our designs[STITT] offered the highest level of safety.


Slide 61

And today, I think we offer the only conventional spark plugs that will prevent the insulators from “blowing-out” of the shell.

For some reason, over the last year or so, both Dresser-Rand and also Cooper Energy Services seem to have excused Champion from furnishing insulator blow-out preventive spark plugs.

So that you can understand the difference, here is a comparison of the blow-out preventive RW80N and the non-blow-out preventive, Y2K version of the RW80N.

GMRC-Slide61Slide 62



Slide 63


Slide 64

You are on a catwalk of a TCVD engine. The engine goes into detonation while you are running your analysis. What kind of spark plugs do you want looking you in the eye?

Spark plugs where the insulators can come exploding out at you with the velocity of a bullet? Or spark plugs where the insulator cannot exit the spark plug shell as if a missile?


Slide 65


Slide 66


I have seen some number of engine modifications made where common sense seems to have been abandoned in the decision-making process that caused the modifications.

In my view, from the spark plug side of the equation, the most cursory visual analysis, preliminary to the engine modification, would have concluded that the plan invited failure and a high potential for risk to life and property.

For example, here is a spark plug that was modified so as to introduce fuel gas to the area of the gap. An operating company bought into this idea and method of spark plug construction.


Slide 67

There is a record of more than several engine fires when this brazed pilot fuel line separated from the spark plug. I would like to think that any knowledgeable analyzer tech would have vetoed the implementation of this hazardous idea. And I would like to think that the decision-making apparatus either did not consider or did override the analyzer techs negative assessment of this highly unsafe idea.

And these considerations and concerns can be applied to other emissions-driven modifications.

Virtually all “screw-in” prechambers have been constructed so as to provoke ignition problems. If not create intrinsically insulator “blow-out” operating conditions.


Slide 68


Slide 69


Slide 70

GMRC-Slide70Slide 71

And these potentialities for catastrophic failure should all have been diagnosed in advance of their exposure to operating conditions.

Lastly, there is the issue as to what will become of even oscilloscope analysis of the high voltage component of ignition system functioning.

Apparently, contemporary operating companies have come to the point of view that ignition analysis is unnecessary.

If this were not the case, why would the operators allow the engine manufacturers to supply them engines that prevent analysis of the secondary side of the ignition system. Since the oil and gas industry would be the largest market for this equipment, what does it say about the influence of engine analysis when engine operating companies obstruct them from doing it? And prevent them from doing it by design?

All of these engines are high speed, four strokes that drive separable compressors. All we can say is that when an operator of such equipment requests our help, we are prevented from being as thorough as we would like to be because the ignition coils are mounted directly to the cylinder head/valve cover, denying all diagnostic tools access to the secondary side of the ignition circuit when the engine is operating.

I would like to recommend that engine operating companies with future engine purchasing plans strongly consider specifying ignition arrangements to engine manufacturers that will offer the operator the means to fully analyze the secondary side of the ignition circuit. I would like to think that this would not be an impossibility.GMRC-Slide71