A talk about API Fluid Loss

• What FL is required and how it can be optimized?

Conceptually, API fluid loss (FL) control is required every time there is a permeable formation. This creates a risk (A: can we quantify this risk?) of liquid phase filtration of the of the cement slurry. That, in consequence, creates a risk (B: can we quantify this risk?) of losing zonal isolation.

A: The answer to the first question is probably yes. It will require Darcy’s equation (where K would be an effective permeability affected by Km and thickness of the mud cake and Ks and thickness of cement cake in addition to damaged formation permeability), delta pressure and viscosity of the filtrate.

Intuitively, every time formation original permeability is in the order of 100’s of mD there is considerable risk of cement slurry filtration; this risk increases for intermediate sections where mud quality and monitoring could be far from desired (not building a mud cake) and when thick permeable zones are present. But, do we care about all this for intermediate sections? You can only honestly care about this if you are going to run cement logs and you are required to record an assessment of zonal isolation if potential flow zones are present.

B: Now, the answer to the second question implies that you are running cement logs and that you are probably interested in a good cement bond/coverage. Lack of FL control would allow excessive filtration and slurry dehydration leading to insufficient cement volume to fill the annular gap and later on radial cracks in the set cement. We see this in cement logs like the micro-annulus (CBL/VDL) and lower than expected acoustic impedance in ultrasonic logs.

Most of the time this is limited to the cement in front of the permeable zone (it correlates to the Gamma-ray and stratigraphic/mud log. Sometimes in previous calipers those zones would appear to be under-gage indicating a thicker mud cake). In some cases, if the permeable zone is long enough a very low fluid loss value could be required ( < 30 – 50ml/30min) due to the considerable amount of liquid phase lost (FL x Length) … Those slurries would be really expensive.

A final word, for any well section, a bad image in a cement log does not necessarily mean lack of zonal isolation, it only means that isolation cannot be attested. The truth is that cement logs more than likely improve with time because nature sometimes helps with this (plugging by flow or creeping formations). But, that would require time, if you plan to do stimulation you could be probably more concerned with the short to middle term.

So, in summary, the FL control required can be estimated by the image on cement logs, taking into account the objective (required zonal isolation, stimulation plans, etc.) Look for your permeable zones and see what the cement log is telling you.

Some recommendations:

–           Start with a high yet low FL control (+/- 250 ml/30min). This can probably be achieved just with cement and dispersant. Sometimes increasing the density 1 or 2 points above neat cement (above 15,8 ppg for class G or 16,2 ppg for class H) helps, e.g., 15,9 or 16 ppg for class G (cement itself is the best FL control additive).

–           Target a thickening time as short as safely possible. Longer-than-necessarily T.T would increase the total liquid filtration. Static fluid loss per se; Static fluid loss happens while the slurry is liquid once it starts setting (> 30BC) filtration is quickly reduced until it stops.

–           For deep wells, more than 5000 meters, a lower FL control could be required to ensure cement slurry stability at bottom hole conditions (FL additive provides viscosity and prevent sedimentation/settling). The BP settling test becomes a reference for required FL additive concentration.

–           Narrow annular gaps and tools like tie back (mule shoe + PBRs restriction/overlapping) would call for a lower-than-apparently-required FL control to prevent cement slurry filtrating on itself.

–           The production zone would require a lower FL to reduce formation invasion by cement filtrate.

• What is the risk of a higher-than-desired FL?. 

As mentioned above, the lack of FL control increases the risk of slurry dehydration leading to damaged cement bonding/coverage. We can see this in cement logs, so if there are no cement logs there is no need to target a tight FL control.

If there is no cement log, then FL control additives wouldn’t be needed unless the slurry, Bottom Hole conditions, rheology, etc. demand a viscosifier/FL product to ensure slurry stability. (I am assuming logs are always run for the production zone, if for some reason, cement logs are not run in the production zone, I would still call for a low FL control cement slurry. (As per the book, to reduce formation damage by cement filtrate). But, if you are going to perforate, cement filtrate damage is much less than mud filtrate invaded zone and it can be bypassed by perforation depth. Anyway, that is something to be discussed with the completions guy)

In practice the, dynamic fluid loss causes significant damage to the cement job by making more difficult to displace the mud in the hole.

This is explained in the following way:

Mud in the hole is composed of:

1. Flowing mud
2. Gelled mud
3. Mud filter cake

The thicker the mud filter cake, the less impermeable it is, so the higher the % of gelled mud or the lower the % of flowing mud. To get the same displacement efficiency the pumping rate required will be higher.

Graphically this can be seen here:

• What is the contribution of mud filter cake?

For intermediate sections, mud cake is likely far from thin and impermeable. In those cases, mud cake should be considered inexistent. Here, most cementers would sell you a train of pre flushes to achieve ‘mechanical’ and ‘chemical’ action to remove the mud filter cake. However, most of the time hardly doing anything.

For production zones, mud cake would probably be closer to its definition. It still needs to be removed, but the task in a friendlier annular gap (smaller) is supposed to be easier and mostly achieved by the action of the slurry itself (talking about abrasive action).

• What is the relationship between filtrate collected and API fluid Loss for cement slurries?

I will use my API 10B 2005 version (API Recommended Practice 10B-2 / ISO 10426-2) to explain this.

The API RP explains when to report FL as Calculated ISO Fluid Loss (tests that ‘blow out’ in less than the 30 min test interval) or ISO Fluid Loss (fluid loss is measured for a full 30 mins without ‘blowing out’).   The FL value is doubled only if there is still fluid filtering at 30 mins if not, the FL is calculated. This is exactly what the procedure says. If at 30 mins, it hasn’t blown out, meaning N2 is not leaking, and it is liquid still dropping, the rate of filtration after 30 mins is not necessarily close to zero, because cement cake is not necessarily impermeable at that time. Now regarding the formula, it is easier to understand looking at the original one:

As you see, if there is still filtrate dropping (without ‘blowing out’ N2) at 30 mins (the end of test). The term in the square root becomes 1 and the filtrate is the double of the collected volume

The formula is my 2005 version is:

Where 10,944 is almost two times the square root of 30

Please if you have any questions or comments, just let me know in the space below.

Cheers

L. Diaz