Cracks in Cargo Tanks – cause analysis and solution

One of our Clients has a problem with cracks in cargo tanks. The problem occurred on other vessels of this type and there is no clue what the reasons of cracking might be. Cracking in cargo tanks can cause leakage of cargo, which is a very dangerous situation, therefore client wants to resolve the problem as soon as possible. The design of the vessel was checked with the shipyard that was responsible for its construction. The shipyard checked all the calculations and didn’t find any clues about what the source of cracking could be – everything looked fine. The client asked class society how to resolve this problem and finally proposed a solution: to get information about cracking from vibration measurements taken inside the cargo tank. Info Marine was the only company that could handle the installation of sensors in such a place. The plan presented by class society was to install 25 sensors inside the tank, 8 sensors on the deck and 4 sensors in the Engine Room. For the whole project, 37 sensor with cables were used. The weight of the parcel with equipment for installation was 240kg.

The problems that we needed to resolve were:

  • appropriate sensors that would not be the source of any spark, even if they would drop away
  • special glue that will maintain its properties for weeks in the tank with cargo fuel
  • special cables that can be immersed in fuel and without damage (the immersion was up to 25 meters)
  • technique of assembly of almost 25 sensors with 4 km of cables inside and outside of the tank without disturbing normal operation of the vessel
  • exit of cables from tank to deck and then to indicated room where measurements were to be collected

We chose special sensors that meet standards EN IEC 60079-0:2018 and EN 60079-11:2012. These sensors together with Zenner barriers can work inside an explosive atmosphere and would not produce a spark caused by an electric problem or hitting a wall due to their properties.

Fig.1 Chosen sensors.

For glueing the sensor, we chose a special glue that has good resistance properties and can be applied in a reasonable time. The whole work was planned on the shipyard and time for installation was limited to reduce the cost to the vessel owner.

Fig2. Glue resistance for fuel in the tank.
Fig3. Time needed to glue the sensors in the tank

Work with glue required heaters to make the whole process faster and smoother. Installation of the sensors inside the tank was performed on a whole wall (25 meters high and 30 meters wide) so the use of scaffoldings was necessary.

Fig4. Inside the tank – installation of sensors.

During installation, it was important to mark sensor locations. We marked all points on the wall and noted the serial number of each sensor assembled. On each end of the cable fixed to the sensor we also put a label with the sensor number.

Fig5. Glued sensor with marked measurement point on the wall.


Fig6. Zenner barrier and cables with labels.

All the cables from the tank were gathered together into one bundle. The shipyard prepared a special construction where the cables were attached to prevent them from moving inside the tank.

Fig. 7 Exit from the tank, the bundle of cables tied to special construction prepared by Shipyard – view from tank.

The most important thing was to secure exit cables to prevent gas leakage from the tank. Special hatch covers and cable glands were used for this.

Fig. 8 Exit from the tank, use of special hatch cover with cable glands.

The whole bundle of cables was secured and routed to the room.


The location of sensors inside the tank – the most important area in our project- was finally installed as on below picture:


Inside the room, we installed Zener barriers and whole collecting equipment.

Vibration measurements were taken in two conditions, with a full cargo tank and on ballast draft. It was very important to exclude the possibility of resonance occurring only in specific conditions.
Analysis of collected data gave good results. At ballast draft load the vibration in the cargo tank was low and the RMS didn’t exceed 15mm/s. Below is an example of the highest observed vibration on sensor number 13 inside the tank:

The high vibrations were observed on the cargo load. We were looking for the spot and RPM of ME where high vibration will occur. As high vibration, We assume 45mm/s in RMS which according to DNVGL rules gives the high possibility of cracks formation. We found 13 spots inside the tank where vibration exceeds the DNVGL limits.    More than half of all sensors (25) are installed in the tank! The highest vibrations were observed on 81,82 and 83 RPM of ME. These RPMs of ME should be avoided till the shipyard repairs of the situation. Below are examples of the spots where the highest vibration was observed:


The source of high vibration was the 6th order of Main Engine RPM. which is correlated to Main Engine firing. The FFT charts ideally show that the main peak is from the 6th order of Maine Engine:


The conclusion of whole project was:
• high vibrations were proven that exist and are the source of crack formation
• high vibrations were observed only in cargo load conditions and on particular ME revolution
• high vibration are the effect of resonance in the frequency range 8,1-8,3Hz
• the force that excites resonance is 6th order of ME

Recommendation for Client:
• Avoid 81-83 ME revolution speed till implementation of the solution
• Install a compensator near the Main Engine that will decrease the influence of the main source of vibration
• Check if tightening/losing top bracings may decrease vibration levels observed in the cargo tank
• Apply stiffness to the bulkhead with cooperation with the shipyard
After applying any solution, the next measurements need to be done to confirm that excessive vibration doesn’t exist anymore.