(69a) Online Sensing of Real-Time Paramagnetic Properties for Flow Assurance Management

As anticipated by Lord Kelvin, an inability to measure a certain component creates challenges and,

perhaps opportunities, for optimization. This has proven true in aspects of the chemical management of

flow assurance in the oilfield, wherein asphaltenic particles precipitate and cause significant fouling and

obstruction to flow. The oilfield industry has developed measurement techniques on aspects of

asphaltenes in solution, but has not previously been able to make online measurements pertinent to the

asphaltene itself.

This changed with the discovery that the paramagnetism of asphaltenic molecules did not depend on

whether those molecules were in solution, flocculated in “near-solution”, or precipitated into larger

constructs. As a result, a real-time measurement of paramagnetism can be used to infer conveyance of

asphaltene upstream of that measurement. In particular, a surface measurement of asphaltene at the

wellhead can be used to optimize a chemical treatment program in the subsurface.

An optimal solution to measuring real-time paramagnetism has been developed based on the recent

concepts of big-data analysis. Instead of measuring the quantity of a certain component once per year,

instead measures quantum chemical properties every minute as the fluid flows past the sensor. Such a

measurement is inherently noisy but AI techniques can be used to extract pertinent features, such as the

concentration of a certain paramagnetic species or hyperfine spacing within the species. One key point is

that the fluid remains at wellhead temperature and pressure and there is no opportunity for oxygen or

other contaminants to degrade the signal.

Such a combination of sensor, real-time digitization and IoT (Internet-of-Things) was developed for flow

assurance applications in 2017 and is now generating data in Abu Dhabi, Gulf-of-Mexico and Central

America. The device works by passing well fluid through a chamber that can be excited with ~4GHz RF

field and a swept DC magnetic field of up to ~1500 Gauss. Unpaired electrons in an orbital will resonate

at a certain ratio of Gauss/GHz where that ratio is representative of the interaction between the electron

and nearby nuclei. In the case of asphaltene, the dominant signal comes from electrons on conjugated

double bonds (Ï€ bonds) found in the aromatic rings and other structural elements. Additional

paramagnetic features can also appear from vanadium, sulfur and iron within the molecule.

The overall spectrum is transmitted by the IoT software to the cloud where machine learning based

techniques are used to extract the π-bond signature. The peak-to-peak value of the signature has been

shown in laboratory testing to directly comparable with the traditional gravimetric methods (SARA, etc.)

that have traditionally been used in the industry. One surprise is that the measurement turns out to be

quite dynamic. For example the asphaltene might alternatively deposit inside a pipe and then slough off,

which gives a sinusoidal variation in the spectral amplitude. Initial results indicate that some wells can

give a characteristic dynamic pattern in the few days before the wellhead completely seizes up due to

deposition – a result that gives significantly earlier warning than can be seen from pressure gauges.

Operators are gaining confidence in the measurement and beginning to incorporate the digital data into

control loops to optimize their chemical management process. It also seems reasonable to infer that realtime

paramagnetic measurements should also have applicability to fouling and deposition within surface

refineries and chemical plants. The combination of quantum chemical paramagnetic sensing, combined

with real-time IOT, cloud storage and machine learning led to the sensor winning the iChemE Global

Research and ADIPEC Best Innovation awards in 2021