Everybody knows why bumping the top of wiper plugs at the end of the displacement is an essential element of each successful primary cement job.
- It ensures only the shoe track is left with cement, and;
- It allows the pressure testing of the casing.
First of all, a couple of questions for you:
What is the reason for a shoe track and how is the length (1, 2 or 3 casing joints?) estimated?
Hint: Don’t be surprised if we find soft cement inside.
Now I am going to detail the most common reasons to explain why top plugs don’t always bump, even after pumping the allowable extra displacement volume.
ID tolerances
The casing dimensions and weights to estimate displacement volumes are typically nominal values as defined by API Specification 5CT. These values have tolerances.
- Casing OD tolerances are +1.0% with an absolute maximum of 0.125” and –0.5%
- Casing Weight tolerances are +6.5% and –3.5%
We derive the ID tolerance from the above. For example: for 9 5/8” casing, there would be up to a 2 bbl difference for each 305m. This value can be greater, depending on casing length, than the typical displacement excess volume (½ the shoe track volume). For example 20 bbl for 9 5/8” casing at 3050m.
This information has no use in the field. However, it is important to realize that casing IDs are different from nominal values. More than likely, the ID will be larger than what is in the tables (or in any calculation sheet/handbook). Consequently, the calculated displacement volume (based on nominal values) would get underestimated.
In the field, this can be avoided by taking casing ID measurements to increase the accuracy of the displacement volume. Typically, the average of 10 random samples for each 100 – 150 casing joints should be good.
Displacement Fluid Compressibility
This is particularly the case when using oil-based mud (OBM), and it derives from the combined effect of pressure (compressing) and temperature (expansion). With compression forces surpassing, the total effect still accounts for a reduction in volume and the need to pump extra barrels to land the top plug successfully.
The drilling fluids company can typically provide the compressibility values for the mud in use; this allows easy calculation of the required extra displacement volume to successfully bumping the plug.
While OBM is undoubtedly compressible, the effect of entrained air can affect both OBM and water-based mud (WBM), mainly when the mud is delivered to the cement unit displacement tanks. Another source of ‘entrained air’ is bacteria growing in the mud.
This ‘additional compressibility’ however can be reasonably compensated in the field by multiplying the displacement volume by the ratio between the density measured with the pressurized mud balance and the density from the atmospheric mud balance.
Rig pumps efficiency
When a rig’s pumps get used for displacement, their volumetric efficiency must be calculated before the cement job to ensure accurate displacement volume.
For triplex pumps the theoretical pump output is calculated from the following equation:
This calculation assumes an unrealistic 100% efficiency. In practice, rig pumps efficiency is anywhere between 90 to 99%, with 95% being an average.
The actual efficiency of a pump gets determined by measuring the suction pit. With this pit isolated, the mud gets pumped out of that pit only for a certain number of pump strokes, then the volume pumped out of the pit is divided by the total number of pump strokes.
Most of the time, the pump efficiency is assumed to be a specific fixed value (written on the board in the company office); however, it can occasionally vary depending on factors like any maintenance done or pending on the triplex pumps or cleanliness of the mud (suction). A safe practice, particularly for the long string, is to estimate the pumps efficiency before the cementing operation.
Most common human errors
- Displacement volume not individually calculated by more than one person on-site for comparison.
- Length errors in the casing or pipe tally.
- In the case of liner cementing, not including the effect of the tool joints (reduction in volume).
- If the cement unit gets used for displacement (smaller jobs); miscounting displacement tanks or not filling/emptying tanks properly.
Other issues
During my early years in eastern Venezuela, there was a lesson learned for deep wells with long 9 5/8” intermediate casings rigorously enforced by the customer: the sole use of Nitrile-based wiper plugs to prevent the effect of temperature and aniline point. This practice was part of the guidelines to avoid wet shoes, including minimizing the rat hole (2 – 3 m Max.); use of self-centralized up-jet float shoes and efficiency of the rig pumps, among other precautions.
The displacement philosophy was simple ‘pump till bump’, but only with nitrile-based wiper plugs. The use of this particular type of material (nitrile) was due to some suspected cases where the displacement fluid (OBM) bypassed the top plug before reaching its landing point. (When its wiping effect disappeared due to wear, temperature and high aniline point (diesel-based OBM) in the mud).
Since then I have rarely seen a company enforcing the ‘pump till bump’ practice. In normal circumstances with a fixed maximum displacement volume (+ ½ shoe track capacity) even if the wiper plug gets damaged and fluid bypasses, the chances of over-displacing (creating a wet shoe) are limited (plug is only left behind).
See the discussion in LinkedIn
Do you have any more causes to explain why wiper plugs don’t bump? Please share
Cheers
L. Diaz
I like this article I enjoyed reading it.
Here is my small contribution to this great article above. Let start with the questions?
What is the reason for a shoe track:
I think the reason of having 1, 2 or 3 casing joints (shoe track) usually left with cement inside, is to ensure that there is good cement on the outside and around the bottom of the casing. The cement is left inside the casing to prevent and mitigate any over displacement risk of the cement (displacement mud volume calculations) and its consequences.
How is the length (1, 2 or 3 casing joints?) estimated?:
Sometimes I think that the choice of the number of casing joints (depending on the drilled section size) is nowadays based on the experience or best practice of WE or Sup Int. However, there is, somehow, a rule of thumb way to estimate it, which is: Required length of shoe track (ft) = Total displacement volume (bbl) x 0.01315 / casing capacity (bbl/ft). Or the simple way is to use an empirical Excell sheet (Film Calculator) to estimate the number of joints needs to avoid any wet shoe due to a mud film thickness removed by the plug. To be considered when facing high viscous mud or lost circulation issue (the well can not be circulated)
Regarding the question why wiper plugs don´t bump?
I think you´ve already mentioned most the reason. I would like to add few things, you may not find them relevant, but:
– First, make sure that bottom and top plugs and float equipment are compatible. Since the bottom plug provides a seat for landing top
plug (end of displacement) and float equipment a seat for the bottom plug, we should ensure their compatibility
– Secondly, high LCM concentration in cement slurry ahead of plugs can be an issue too.
Hi Bikouyi,
Yes, the shoe track ensures good cement is at the shoe and around. Its length depends on the length of the casing and the rule of thumb you mention is indeed used to estimate it, however there are other considerations, like the planed completion for example. You can also play with the thickness of the mud layer and make your own estimation, if you want to be more detailed and not use the rule of thumb.
The couple of additions to answer why the plugs don’t bump are also very applicable.
A quick one what does 0.01315 represent or where it come from, is there a way to calculate it?
With the introduction of multi-stage fracturing, and more specifically, a hydraulic “toe-sleeve” for flow initiation, it has become imperative that the pump-till-you-bump method is practiced. If any cement is left in the casing, the results will be very costly.
I have seen a few cementing companies “wash-up” on top the plug, or even put their pump-out tee off the back of the truck (rather than off the plug manifold), resulting in just a few gallons of cement ending up in the toe sleeve, rendering the sleeve useless. (Could also happen from diminished wiping efficiency of the plug). Generally, a gun run needs to be completed in order to create an initial flow path but in a case where there are diminishing seat sizes (frac sleeves/ports), those few gallons of cement can be catastrophic as most of the seats need to be milled out to accommodate perf guns.
Further to that, there are actually combination float and plug systems that are designed to achieve an engineered “wet-toe”, which is supposed to help preliminary frac-initiation subsequent to the toe-sleeve opening event, all while allowing a plug bump indication.
Both scenarios require a bump, both to ensure proper cement placement, but also to accommodate regulatory requirements (casing test prior to frac).
As for why they wouldn’t bump, there have been instances where the plug fins either couldn’t collapse enough, or the plug core was too large to fit through the above-mentioned frac seats causing a premature opening event due to increased piston area / pressure drop and therefore force to shear the frac sleeve to the open position.
Very Interesting contribution Mr. Clayton. Thanks a lot for this.
Casing Properties
I have looked at the excellent post by Lenin on bumping the plug during cementing operations and the many difficulties that can be encountered.
I am, however, very disappointed that there has been no feedback from the many engineers that follow our page.
I expected more.
Some people seem to find it easy to send me personal messages on messenger and expect immediate answers on the same subject. My preference is always to share knowledge publically rather than share it with a few people.
Please share your comments on this page so others may learn from you.
Better to light one candle that shows the way, than just curse the darkness.
Casing Properties and Bumping the Plug: A Tiny Candle to Light Your Way
For critical well, you must understand fully the tolerances allowed by API.
My advice to all operations engineers is to drift (all) and measure (ID, t) of at least 20% of your casing and tubing strings to find a true average of ID, OD and thickness.
This will help you with cementing, monitoring casing and tubing wear, casing strengths, etc. Task is not easy!
To appreciate your task, just look below at the specs we use in the oil industry and how confusing they can be.
Inside diameter ID is given by
outside diameter minus twice the “nominal” wall thickness.
ID = OD – 2 x t
Outside Diameter
From API 5CT, OD tolerance for pipe smaller than 4.5″ is +/- 0.031″ .
For 4.5″ or larger, OD tolerance is +1.00%, -0.5%.
Wall Thickness
The actual wall thickness of casing can be 12.5% less than the “nominal” size and is still within API limits.
From API 5CT, tolerance on wall thickness is 0 -12.5%.
The “nominal” wall thickness is the maximum wall thickness used to calculate the casing strength properties.
Nominal wall thickness is provided by manufacturers.
The tolerances on wall thickness are much more important for casing strengths than the OD.
The wall thickness cannot be greater than nominal – it can only be less i.e. thinner.
EXAMPLE: Given 10,000 ft of 9 5/8” casing 47 lb/ft with the float collar positioned at 9,820 feet.
In wells using high density muds , it is possible for the inside of the casing to have a thick film of mud, which can be as much as 1/16”.
The total volume of mud stuck to the inside of the casing is given by:
ID of 9 5/8” casing = 8.681”
Effective ID of casing = 8.681 -2/16 = 8.556”
Volume of mud film = [ ( ID^2 casing – ID^2 effective )/1029.4 ]x length of casing
= [ ( 8.681^2 – 8.556^2)/1029 .4 ] x ( 9,820) = 20.6 bbl
Two problems will arise
1. If you do not use a bottom plug ( like some new engineers are currently advising their companies to drop the use of bottom plugs!), then you will have a wet shoe!
2. If your ID is not 8.556” because your tolerance is not calculated accurately, then your displacement volume can not be calculated accurately. You will never know if the plug bumps or not.
I have seen a few of those.
Remember: the rule is not to pump more than half the volume of the shoe track if you do not observe a bump.
The above example clearly shows that the volume of dirt that accumulates on inside the casing can be more than the volume of the shoe track.
In this case the plug will not bump based on the calculate displacement volume. This error arose entirely from using the wrong value of ID.
Now you read this post, tells us:
1. Do you you regularly observe plug bumping in cementing operations? Show % if you can.
2. Which casing size bumps easier?
3. Which ID do you use and where do you get the value from?
4. Anything else useful.
I will elaborate more on the above once I see reaction to this post and Lenin’s post.
Regards
Hussain Rabia, Dr.
Thanks Dr Rabia for this excellent and valuable contribution.
Dr Rabia, many thanks for sharing this information, coming from the drilling side there is a lot of useful information and rules of thumb which i will find very advantageous !
Thanks for your contribution Mike.
Cheers
L. Diaz
Hello All,
I have a query about bottom wiper plugs. The plugs when tested at in-house give accurate rupture pressures per design. These plugs are tested at same temperatures as that of BHCT per customer’s requirement in HPHT cell. But during the actual job they show high rupture pressures (200% more). Can someone share you r experience or opinion , what might be the probable causes here? (they have rubber diaphragm)
Keshav
Thanks for your contribution Keshav.
This is a good question, a question that is actually more common that it should be. Almost in every operation there are always cases of bottom plugs not bursting or showing higher-than-designed rupture pressures.
And in every single operation where this happens there are always theories and explanations, which are never proved.
The best thing we can do is to mention “probable causes” and make an action plan that is assumed either to work or to reduce the possibility of another even to happen.
However, I remember a case, where a customer tired of never finding or given a solution, actually POOH a casing full of cement after a bottom plug did not burst. The supplier always assured with evidence that plugs from the same batch where all tested and re-tested to rupture as designed.
After the casing was POOH the shoe track was cut and the bottom plug removed intact with the diaphragm also intact with no signs of any cutting, debris or other material below or in the diaphragm chamber causing any apparent obstruction. Then the plug was tested and it required twice the design pressure to rupture.
In my opinion, I have had this situation several times. In another case, We had far to often cases of bottom plugs not bursting. Probably in 20% of the wells (9 5/8” section), this was a batch-drilling operation, meaning all 12 /1/4” sections where drilled in batch mode. The customer asked to change the casing hardware supplier and we did. Then for everybody’s surprise the problem continue with exact the same frequency!!. In this case, we finally managed to prove that using water, as a pre-flush was the reason. When we change to a weighted spacer the problem solved.
Now going back to what wiper plugs do, it is clear that there is always a possibility that the nature of the mud/casing wall interaction, with or without the help of the pre-flush, the bottom plug carries underneath excess gelled mud wiped from the casing wall. If this happens, then it comes to the design of the plug and float collar (landing point) to manage that situation. Ideally, all this should go thru the float equipment. This possibility in increased by the presence of intended or residual LCM, cuttings or even a flocculated mud system.
Then in comes to QA/QC of the casing hardware (wiper plugs) and we all now the supplier will always prove or find a way to prove this is perfectly done.
Anyway, I believe the root cause for some of the problems is basically a weak point in the process: the time and conditions casing hardware is keep in store. That can have a tremendous effect on the performance of the wiper plugs. However, it will remain random and difficult to track.
Something, I can say about all these events, it is that they are more common, around 75 to 90% of them (not including well with liners), in intermediate sections or deep surface sections. These are the sections where drilling fluid QA/QC is not best, more drilling events (influx, gas, losses) tend to happen, etc. Maybe that has some contribution to the occurrence of these events.
If you are witnessing higher rupture pressure, up to 3 times the designed value, it is just a matter of time that an NPT event would happen. I would recommend having a look to the things I mention. Remember: Mud and storage conditions.
Hope this helps
Cheers
L. Diaz
Thank you so much Lenin for the detailed explanation of the possible causes..!
Very interesting discussion , thanks everybody. I had several times in 9-5/8 csg @ 1000m when using two plugs that after bump a top plug the shoe was without cement. As per investigation the job was executed perfectly except the job was done with total losses and drill under water with total losses, the well fluid is water. The requirement was just to have several meters of cement to isolate shoe from next section 8.5 “‘. Thus this practice using at some areas still . One of the reason we think that due to free fall of cement it might create air gap under top plug witch can occupied a full shoe track volume (it is one joint only). Or common conclusion from customer the top plug is failed and (displacement) water just bypassed. Anybody support this idea that possible “get air / vacuum” below the top plug when it is bumped, or just suck fluid above top plug even the top plug is OK, i.e. not damaged.
? The 9-5/8″, ID 47# of the casing was very 8.76-9.00″ and used HWE plugs (non rotating).
Thanks!
Alex thanks for your very interesting question,
This is the typical situation in our industry where there is uncertainty about the root cause of an event due to the difficulty to validate most hypothesis. However, I don’t believe the theory of the air gap underneath the top plug due to free fall. Here we have to understand that the vacuum can only exist near surface dynamically. Depending on the timing to drop the top plug and start displacement, some air might indeed come between the cement and the top plug. This air in normal conditions is quickly compressed and/or displaced behind the top plug (suck by the vacuum effect that exist behind the top plug) and inside the displacement fluid.
The effectiveness of the seal between the wiper plugs rubber and the casing inner surface is not perfect, so air leaking from any air trapped underneath the plug will happen facilitated by pressure differential.
Now talking more about this dynamic seal, it is definitely more effective with weighted mud (able to leave some sticky mud residues adhered to the casing walls which is wiped/removed by the wiper plug and promote sealing). With water, assuming water is your control fluid and not mud. The following can happen:
-If water was used after the initial mud volume was in the well, due to the losses scenario you mention, then any sticky mud residues were, at least partially, cleaned by shear
-Pumping rates are likely higher or high enough to allow water to leak behind the bottom plug into the cement slurries
-Wear of the plug rubber outer surface (fins) in contact with the casing is increased, which also facilitate water-leaking pass the bottom plug from below
-Water density facilities further all this and will quickly float inside the cement slurries. Then depending on the length of the casing, the well deviation, the rheology of the slurries and density, the water can quickly reach the top plug or at least contaminate (dilution) the tail slurry.
Finally, there is something obvious; the wiper plugs must be adequate for the casing size and weight. Which means that there is interference between the plug OD and the casing ID. If the wiper plug OD is less than the casing ID (lower weight, including ID tolerances from the manufacturing process). The possibility of leaking water for below the bottom plug is much higher.
In summary, I believe water is leaking from below the bottom plug and migrating to underneath the top plug for the reasons explained above. Now using Stoke’s law We can probale estimate the velocity (time) required for water to move up (float) in the cement slurries … to answer the question: Can it move fast enough to get to the top plug? … (remember the lenght between plugs also shorten during the job) For this, if you are interested, let me know and contact me personally … I can probably give it some more thought to this.
If all what I said is the most-likely root cause for your shoe track without cement, then you might have probably found hard cement (could be harder than what it should) somewhere below the shoe.
Hope this helps
Let me know
Cheers
L. Diaz
Really a very productive discussion and and an excellent platform I have come across!
I am Well Integrity Engineer and in many Wells we need a perfect circumferential Cement bond above the Casing shoe. Due to losses or other reasons the same can’t be achieved.
My question , can we increase the Shoe track length upto 5 or more joints or more than 300 feet for a better Cement across the Casing shoe??
You also have to consider when making up the CMT plugs into the CSG HGRS, a lot of the assemblies are made up onshore and sit for weeks horizontally. While the assemblies are lying horizontal the plugs are subjected to side loading which could effect their performance when run, as one area of the plug set is deformed.
Hi Baz, very interesting point you are making.
Thanks for your contribution, is there a potential or practical solution to this problem?
Cheers
L. Diaz
All,
I have not seen aerated mud discussed as a cause of failing to bump, but this has happened to me twice recently. I appreciate any comments or suggestions on the following “non-routine” proposal to make sure I’m not missing anything.
On a jack-up well recently drilled in SE Asia, we under-displaced two primary casing cement jobs (13.3/8in and 9.5/8in) due to aerated WBM. The mud system used was a “high performance” inhibitive KCl/polymer system with an amine-based shale inhibitor, 9% KCl, and ~56k Cl-. It is notorious for this aeration, although the severity is inconsistent and unpredictable.
Back-calculated pump efficiency (compared to the theoretical displacement based on micrometered ID’s) was 94.3 and 95.7% using the rig pumps and the cement pump, respectively. Both jobs used full-bore, non-rotating, PDC-drillable, double plug sets launched from a surface-release, double-plug head.
In addition to more conventional measures, I am considering dropping an extra bottom plug when we first land the casing and do the pre-cementing circulation. This would give us an accurate strokes-to-bump figure immediately prior to cementing. It would also require a bit of extra time to load the full plug set into the double-plug head once the pre-cementing circulation is complete, and then to drill through an extra bottom plug (with a PDC bit). But that time would still be less than the NPT incurred drilling out excess cement, and would be time well spent in achieving proper cement displacement.
The general sequence would be as listed below. (Apologies if “too much information”.)
1. Offline, load two bottom plugs into the double-plug cement head.
2. Run and land casing. Make up double-plug cement head on top of string.
3. Rig up and test cementing line.
4. Back out bottom plunger to launch first bottom plug.
5. Displace casing volume with rig pump, noting strokes to bump and burst the bottom plug burst disc.
6. Complete pre-cementing circulation. Shut down pumps.
7. Run in lowermost plunger for bottom plug.
8. Remove double-plug cement head cap.
9. Retract upper plunger for second bottom plug (initially installed in the top plug position); push it down into the normal bottom plug position.
10. Run in upper plunger for top plug.
11. Load top plug, re-install cementing head cap.
12. Cement casing as per normal.
Thanks in advance for any feedback.
Hi Paul,
Thanks for your contribution.
An additional calculation you can do is to multiply the displacement volume by the ratio between the density measured with the pressurized mud balance and the density from the atmospheric mud balance, as a reference against your actual measurement with the extra wiper plug. Please let us know the outcome.
Cheers
L. Diaz
Appreciate the insight from the experts on this topic – very interesting!
Bit of a different topic but, I was wondering what everyone’s thoughts are regarding decreasing rate during cementing prior to landing plugs? I have seen a lot of wells where coil is deployed into the hole and “cement stringers” are encountered close to the toe. Is it possible when rate is decreased cement bypasses the plug? If rate change contributes to the likelihood of leaving cement behind, and if anyone knows of any papers around this, I would be interested in learning more on this.
Hi Ryan, that is a very interesting question. I am curious in knowing more why you suspect this could be the case. Certainly, I can see a relationship with the pressure applied on the plug and fins as it is being pushed down the casing, perhaps at some point the leyer of gelled mud on the inner casing walls starts to open paths for the cement to flow through ahead of the plug, particularly a lower rates. I don’t know.
Honestly, i heard this before and I think even thought about it as a possibility for cement-left-inside-casing events.
Definetely interested in hearing more about this and will look for information, If I come with anything giving clues I will share it here
Cheers