Poll Results: To Separate or Not-To-Separate, is this “Still” THE question?

To Separate or Not-To-Separate, is this “Still” THE question?

Making a poll in 2025 about Separators versus MPFMs can sound like a joke for a few of you and a serious question for others, but is it the right question? Or is it only the starter of a conversation for THE REAL questions?

This is not a debate on new versus old, early adopters versus laggers. Those who favor only Separators look at MPFMs as this black box who gives numbers that can’t be checked physically; Those who favor only MPFMs have dealt with a leaky old separator with a doubtful Turbine meter or PDM,  a dirty oil/water interface, and believe this old thing should be out phased.

At PandA, most of us pride ourselves on possessing extensive experience with MPFMs. Given that, would our response to the conclusion of this poll be PRO-MPFMs? This must be the expectation of most of you.

The fact is, in PandA, we are not Pro-MPFMs or any measurement devices. We are Pro-continuous accuracy in measurements.

To answer the question, let’s use a car dealership analogy: YOU come to us looking for the best, most performing vehicle on the market, and we point YOU towards the Ferrari of measurement devices.

Please don’t expect that. Our answer will always be based on our motto:

Integrity, impartiality, transparency, and unbiased guidance.

First, we ask YOU questions:

• What is your current car? Why would you like to change it?
• What is your usual trajectory, or driving routine?
• What is the one component you would like to improve against anything else (speed, consumption, safety, reliability, efficiency, maintenance, ease of usage…)?
• Where is your car going to be parked and used?
• What is your budget?
• And finally, the question that most car dealerships will never ask you: Do you really need to change your car, or would it be simply enough to enhance some of its features?

You might want a Ferrari, a beautiful, shiny, latest, pricy model, but do you really need it? Is it suitable for You?

In Flow Measurements, there is no such thing as the “Ferrari” of measurements. Whatever Manufacturers are trying to sell you (and most of them are…), there is simply the best measurement device, or optimised solution for YOUR specific cases.

In other terms:

• What is the fluid being measured?
• Where is your measurement device? Or where would you like it to be?
• What are your current installation and its pertinent issues?
• What are the flow conditions?
• What is your budget?
• What are the vendors/manufacturers in your region who can provide high-level maintenance support?
• Is your team comfortable working with the most advanced equipment?
• Do you really need a new device or only a new system/protocol to improve your current measurement?

This is the first thing to have in mind, is it : Heavy Oil? Black oil? Volatile oil? Condensate? Does it have water injection? Steam? Gas injection? Chemical injection? Composition changes?

Any flowmeter, either single phase or multiphase, will provide a measurement at line conditions (pressure and temperature) or also known as “observed conditions”/“meter conditions”. What we need to report in terms of production is at standard conditions, meaning we will need to understand its fluid behavior (aka PVT, EOS…) to do this conversion. This path will not be solved by any flowmeter. You will need to combine the measurement with a conversion tool/PVT .

We have seen far too often companies using Black Oil Correlations while  working with a wet gas or using outside of the domain of application of developed correlations. If you are using the black oil model for condensate (for instance) the CGR (Condensate to Gas Ratio) will be wrong, are you using ALL the right parameters for your flow calculations?

Now, let’s look at fluids beyond the hydrocarbon range: sand production for example, makes it challenging for both MPFMs and Separators, but then the separation itself allows for sand collection and avoid erosion downstream the line. Another example is high salinity or chemicals which might taper with not only attenuations-based measurements, but also permittivity-based water cut measurements.

Using the previous analogy, it is the same as driving a Ferrari in the jungle and wondering why it does not work as advertised.

 

Figure 1: Flow path from a measurement device to standard conditions for oil, gas, water, and solid

 

2.    Where is your measurement device installed? Or where would you like it to be?

Flowmeter placement is a crucial element affecting the fluid behavior modeling, which depends directly on pressure and temperature, closer or farther from the wellhead the measurement is made (or expected to be made), and more impacting will be the PVT package used.

Let’s start with the line pressure considerations: The largest deployment of the MPFM fleet is rated between 3 000 and 5 000 psia (with some at 20 kpsia); The largest deployment of Separators handles 1 440 to 2 000 psia, but some can handle up to 15 kpsia as well (API 6A class separators). This generic hardware statement usually points at higher pressures and larger flowrates to be handled with MPFMs, and low flow wells and late-life fields to be handled with Separators.

Accessibility is an important factor as well: a Separator should not be used in remote places because of the difficulty/safety/unlikelihood of full remote automation, for example: Subsea installations prefer largely MPFMs to separators, since they are relatively easier to handle, and safer. Yes, there are also a few 4-phase Separators designed for ultra-deep water, are they necessary? Sometimes, yes. Are they challenging to work with? Absolutely!

Location beyond placement, might also narrow down some MPFMs deployment due to radioactive sources regulations within countries, or company policies.

A few years ago, we worked with a company that, after our consultation, redesigned their layout by simply relocating their MPFMs at the manifold, and closer to each wellhead. This eliminated the need for a shared mobile test Separator, which had only been used intermittently (once a month per well, then using those rates as average production). This change enabled true flow optimization thanks to the real time readings with an evolving production (the consistent choke changes, well shut-downs, re-openings, tie-ins) and allowed for comprehensive well performance mapping, and ultimately to measurable production gains.

 

3. What is your current installation? And what are the pertinent issues?

Companies look at the measurement uncertainty (often miscalled ‘accuracy’), as manufacturers promote this type of differentiation by giving the lowest uncertainty possible (while playing with confidence levels or generalizing the sweet spot of the MPFM). However, in our market review and collected information, most of the installations (over 50%) did not call for a low uncertainty target, but a repeatable and reproducible measurement. A perfect example will be optimization (i.e. gas lift…). In such a case, it is the trend that is important, and the difference that you gain or lose versus the previous conditions (i.e. by injecting or reducing the gas lift).

A repeatable meter has a given cost, a reproducible meter is more expensive, and a low-uncertainty meter is way more expensive. Following our analogy with the car industry, select the right car for your road conditions, this is regardless of using an MPFM or Separator.

AKA, one of the reasons why a Honda Civic or a Toyota Corolla might sound boring or slow next to a Ferrari, but in most cases there is no need for it regardless of the available budget.

 

4. What are the flow types or flow regimes? (stable, intermittent…) And flow conditions?

Given the design of a Separator, the interface liquid/gas has to be set (level of liquid in the Separator) which smooths the flow to a more stable line and better average regardless of the flow types upstream. This could be better in some cases and detrimental to well evaluation in others.

The definition (i.e. frequency acquisition) of the flow production is the point to consider here. You will see all variations in real time with an MPFM, less with a Coriolis-based Separator, and close to none with a conventional Separator. If the selected MPFM has a fast acquisition system (> 10 Hz for flow rates reporting), you will see the entire dynamic of the flow conditions. This is akin to having the heartbeat of each well. It is the RPM engine of your car. You can now select if you want to run faster (open choke) or not.

Keep in mind that most of the test Separators assume a perfect separation, they do not guarantee it! The essential part forgotten in our ring engineewording is:“gravitational test separator”; this means we work with the gravity principle, and if we have a lower density contrast (high pressure, heavier oil, viscous fluid…) or one phase significantly dominant (wet gas), then more challenging will be the “accurate” flow measurement. If the well is “slugging”, then we have another dimension to the true measurement.

 

5. What is your budget?

When it comes to the money aspect of it, two things are often confused: Price versus Cost. In general, most Separators are relatively cheaper than most MPFMs, again, this is not written in stone; an example is the Subsea Separators, which make Subsea MPFMs look relatively a bargain in comparison. This is only the price of purchase, beware of the not-so-well-hidden costs! We have a joke at PandA that there are two teams within each large purchase project: the one who performs the sale/purchase based on the benefits(or price), and the teams who pay the real price; AKA: those who decide and those who use and maintain in the aftermath (struggle with wrong devices).

For Permanent installations, an MPFM is a pricey acquisition, but being a tube with low interference and no moving parts, it does not carry the same scheduled upkeep and verifications that a Separator needs, it does not expect a permanent team weekly checking the valves and levels, this is the cost we talk about. Also, when it comes to cost vs exposure, the worst that could happen with an MPFM is a bad measurement, the worst that can happen with a separator is an environmental disaster…

Even when it comes to MPFMs, the most popular or expensive one is not always the best solution. Sometimes, as said earlier, by selecting a few relevant flowmeters set at the right location, we can significantly reduce the cost of production monitoring.

You should not be surprised that we may recommend a very simple solution or a conventional separator sometimes when the experienced team is already in place, as one of our advisors says: “A separator is an MPFM which is only 160 years old”.

Estimating the entire expenditure is where you need to be careful as an end-user. PandA looks at the whole scope of the deployment of your solution, not only CAPEX but also OPEX. MPFM hardware is becoming cheaper with newcomers and their specific solutions, and a very large background from previous experience. This is going in the sense of more deployment of such technology, but keep in mind what the yearly OPEX is expected to be. Manufacturers may try to work towards this solution by reducing the CAPEX and expecting to make money on long run. (This is also called the Elevator sales model: cheap purchase, outrageous maintenance)

At the end of the choice, companies could be interested not in buying but in leasing as a solution. This can be very cost-effective, but a proper evaluation should be made versus time, and this should be linked with how much you wish to know and take ownership of the flowmeter technology that you are using or possibly buying. Does investing in a car make sense? Not always, leasing allows for clear services and time retrained contracts. 

You may sometimes combine a few technologies and get a hybrid version way better than what is on the market. Manufacturers will probably discourage you, but there is no perfect solution on the market that solves all of the problems 100%.  

A dedicated approach to your problem should be taken and examined to see which benefits you could get from this approach. This is what we as PandA specialise in.

 

6. What are the available vendors in your region who can provide high-level maintenance support?

When it comes to maintenance and general support, not only are all manufacturers not alike, but the response quality and speed of support within the same manufacturer differ completely from one region to another. This is an often assumed and neglected aspect of purchasing new metering equipment. In the long term, this is crucial; if your device is not set according to specifications, needs a sensor change, or maintenance, and it takes hundreds of emails and many months to get support, no matter the technology complexity (or lack of), it is a headache.

Any system has a given reliability, and the more sensors a Separator or MPFM has, and lower the reliability. The more openings (fittings) in the main pipe you have for sensors, and the more you are exposed to a possible failure in the pipe’s integrity. The need for reliability and safety is high, and this is driven by the policy in your company, and this may lead to some solutions being more popular than others..

Any rotating equipment (turbine, positive displacement meter…, etc.) and any configuration without a full-bore solution will or could cause issues (e.g., ageing quickly, or changing locally line pressure and temperature leading to hydrates or other types of depositions).

The quality of the fluid, or the presence or absence of slugs, could sometimes be detrimental. Sure, the separator is not the best choice sometimes with moving parts (valve opening and closing) to keep an interface gas-liquid leveled in a separator and expect the best separation, but if you are in very stable flow conditions, then a Separator does the job perfectly.

Any flowmeter failures can be catastrophic sometimes. Imagine a subsea flowmeter not detecting the water coming, and then there is hydrate formation plugging the pipe. Imagine water coming in a well with H2S gas, then this could call for some specific material, procedures… If you have a very high temperature and pressure, some exotic material may not be suitable for the long run. If you have a large change in pressure and temperature, this could add stress and ageing to the equipment.

Leaving the installation with the metering experts is what we believe is essential, and the manufacturer, with the integrator (EPC), can do a great job. You may have some warranty become obsolete if this is not done correctly, or you may face issues if you want to claim that the meter is not working. Meanwhile, this should be managed by the Manufacturer’s support team, and this team should be able to intervene relatively quickly and be there when an issue arises and get the system update. You want continuity in services. The purchase of a flowmeter leads to decades of use of this technology.

You want to work with a team who can say “yes, there is a problem”, “yes, this is our fault”, and then we fix it. You want to be sure you are up to date with the versioning of software, there is always improvement, and take advantage of that. You should have, as a minimum, a free training available for your team included in your contract, and reccuring at each update.

 

7. Is your team comfortable working with the most advanced equipment?

Installing new technology might be thrilling for your local team, showcase the company, and teach a lot, but is the training provided truly cover the new parameters to be measured to set the MPFM, and how to keep it running? Are some health check indicators present to know when something is wrong? Do you have the capability to interact with the device by yourself? Nobody knows more than you about field operations, when for most of the MPFM manufacturers, their field experience is limited to a few people. They are experts in metering, but not necessarily in field operations. The gap between theory and practice has led to PandA, in many situations, needing such interventions and mediations between companies and manufacturers with obvious knowledge transfer.

Are you willing to spend time learning? Are you able to keep this field in production this way for many years? Some countries have developed incentives (training, KT) to have the knowledge passed to end users, and overall, this is paying back; but having little control and limited knowledge over what is used is something we often see and have to remedy to .

We believe they should always be a preference for manufacturers willing to be engaged and share the basic information and principles that your instrumentation team can handle daily when having minor issues. You want to keep the uptime of any flow metering device as high as possible.

 

8. Do you really need a new device or only a new system/protocol to improve your current measurement?

Changing equipment or getting a new product is a large financial investment, but also in time, learning curve, and deployment. Before going with a new solution, it is important to understand the actual problem first, and the real reason behind the existing solution’s inadequacy. Getting a new product does not immediately solve all the potential issues, and will bring on new challenges. Have you checked some typical parameters like repeatability or reproducibility to understand why the current solution is failing? Did you check that the possible discrepancy was because of wrong PVT inputs? Was the recording long enough to capture the production signature of the well? Is it much more of an issue with optimizing or a maintenance program?

We have worked with companies deciding to move to new products, but the actual need was a clear awareness of the effort to spend regularly to maintain the current basic flowmeters.(AKA: it’s not a meter issue)

We have helped companies that tried to compare technologies, but failed to use the proper process to evaluate the performance of one type of measurement against another.

We have consulted with companies struggling to see their real production, thinking that the solution would be the purchase of new metering devices. Yet the solution was simply to relocate some existing meters, update the PVT for some ageing wells, and establish a new testing protocol/schedule to get a more accurate mapping of their production. Only then were they able to optimize it.

 

Conclusion of the survey

The poll was created to show the split existing in the industry about how experience shapes beliefs, and none of the poll options were wrong, or truly right  for that matter. Context determines solutions, and each context is unique, which makes each solution unique.

However, we also need to break a dogma and a very misleading assumption that Separators do not require PVT or fluid behavior, which is why they are easier to use. None of the flowmeters anywhere in the industry can anticipate what the process will be downstream of the observed conditions (metering conditions) because we are dealing with multiple phases always at a bubble point or dew point. We need to use these fluid properties (aka conversion from observed conditions to standard conditions). The fluid has an essential role in the performance (i.e. uncertainty contribution) of any flow metering device. As one of our experts in PandA demonstrated to a client: the key point is to look at the overall uncertainty of your complete metering solution, one of the parameters you can fix immediately is the input data (setting of the MPFM/Separator), and the proper maintenance and conditioning of the meter. Let’s take a couple of examples:

1. If you have wet gas flow passing through an orifice because of poor gas separation, the gas flow rate could be 30 % or more off with a separator.
2. If you use the wrong water and oil density information with a Coriolis to calculate the WLR in a liquid line, the data will be way off.
3. If the fluid composition changes (example: water is produced or salinity changes), most of the MPFMs will be affected through the fractions measurement and report wrong Gas/Oil/Water production without additional information.

 

Note: If you would like to discuss this article, or consult PandA for your next car dealership (or measurement device and its uncertainty), do not hesitate to get in touch with us by email: (enquiries.panda@pandassociates.consulting), on LinkedIn, or at the contact below!

© 2025 P and Associates Sdn. Bhd. 202001010646 (1366966-D) All Rights Reserved