Portal Industrial Cartagena Colombia - Forum - Contacts

Members Login
Username 
 
Password 
    Remember Me  
Post Info TOPIC: level sensor diving operation


Guru

Status: Offline
Posts: 12637
Date:
level sensor diving operation
Permalink   
 


level sensor diving operation

 

Project info recieved :



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

Good morning Wilfried.



I know it is short notice.

We urgently need an electrician (and spares) for the .... Vessel will arrive today, departure tomorrow.



The problem is the high level sensors. Please see the attachment.

Inside the stem are 2 reed switches with a resistance. We do not need the complete stem, only these reed switches.

Does not need to be the original ones. Might only be the resistance.

Info from Chief engineer:

Today we removed last remaining magnetic level switch from Stbd Slop tank and installed inside stem on 1Port Cargo tank. Adjusted and tested and was working OK. After lunch test have been repeated and switch was out of order. Also have been found that high level 95% alarm on 5Port Cargo tank lately repaired is again out of order. Tested with Fluke instrument and found that both reed switches when activated by floating magnet shows ca 5K and there is no alarm. Alarm is activated when reading is above 10K like on 98% alarm. At present on 1Port and 5Port Cargo tanks we have only 98% overfill alarms and on Stbd Slop tank both alarms are out of order.



(Maker: INELTEH d.o.o.

Ljubljanska cesta 7,

51000 Rijeka, Croatia)





Diver inspection:

The vessel touched with the bottom in Paramaribo.

We need a GL approved diving company to make an under water inspection.

(Ordered it for Santa Marta, but the diving company was rejected by the surveyor due to the missing approval).



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

 



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

Project status 12:53

Diveteam in standby for 6am tomorrow anchor site Mammonal.

 

Access window according to ship shedule arranged by Agent .



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 


info update 14:06 LT


Good day,

There are two different jobs :

1. Repair of High Level Sensors. This is job is for electrician and can be started any time if permitted by Terminal.

2. Bottom inspection by diver (must be GL - DNV approved company) to check ship's bottom after touch the ground in one of the previous port. This job to be done only when vessel is on anchorage.

Ship's schedule:

12/1100 to 13/0600 at Oil Tanking Berth.
13/0600 to 13/1200 on inner anchorage.



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 



It is the high level alarms in the cargo tanks 95% and 98%. On open deck

The diver inspection is only to chech the hull for possible damage due to touching the bottom during discharge. Here we need the GL Class to sttend. Will call the surveyor when you confirm the diver attendance.



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

.

 

Dive inspection in nightdive



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

klaus said:

Hello Wilfried

High level alarm
2 pcs resistance,
1 Kohm in series and 10 Kohm in parallel of reed switch,

1 Kohm when contact closed and 11 Kohm when contact open

__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

Wil said:

Yes, we understand that - but we built 23 versions of the sensor with all kind of combinations of reed switches and slight resistance variations - and could still not get the circuit board accept the stem signal as valid signal - only 2 Inelteh fabricated sensors mounted precisly give a "accepted peripherical" signal - we are still unclear why that is so - intention from Inelteh is suspected...



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

Top view sensor stem



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

buoy sistem



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

sensor original broken with reed switch and resistance resin cast by inelthe...



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

Fabricating sensors from reed switches...



__________________


Guru

Status: Offline
Posts: 12637
Date:
Permalink   
 

CARTAGENA REPORT

In the sensor we deal basicly with a mechanical part (the float buoy and its rail) and a electonic part that detects the presence of the buoy and triggers the alarm in 2 precalibrated positions. When confronted with sistem failure the crew has to diferentiate if the malfunction is mechanical (the buoy does not aproximate the sensor is jammed etc..) or the buoy is working mechanicly but the sensor doesn't detect its presence properly. A good way of testing the funktion of the sensors (and diferentiate between mechanical buoy failure and electronic detection sistem failure) is having the detection stem on deck and triggering the levels one by one with a magnet. Without touching the buoy sistem.
What the crew finds when opening the deckplate is a 4 wires going down the tube identified as "sensor stem".

1) recommendation

Holding those sensor stems (the tube with the height calibrated detection assembly) as original spare part from Ineleth will be a good solution for a straigh forward repair the crew can perform every time the sensor sistem fails.
The testing if the sensor stem actually is the problem is also straight forward with a magnet looking if the proximity of the magnet fires off the alarm on the alarm table.

As this "spare part (the whole stem from Inelteh) " was not available this time, and we had the correct technic on the set (he is designig and building fill status detection sistems for the port and for the refinery industry) we tried to get a handle on the sistem by "reverse engineer it" understand it and reasemble it - what got us to the repair of 2 of the failing 4 tanks.

It needs to be said that this kind of repair is far beyond what a normal electrician can do and understands and it is certainly beyond "spare part replacement" and "getting a standard spare part" on the market.

Pulling out the 4 cables that go down into the tube identified as sensor stem you find a detector assembly for each level - the detector contains a normal looking reed switch and a resistance cast into retsin by Ineleth.

So we had the cables and 2 detectors pulled out and could fire the alarm with a magnet on deck without a problem, but when the cable tree with detectors was pushed gently down the stem tube obviously the distance of the dedectors is altererd a tenth of a millimeter and the sistem detects this minimum distance alteration between the two detectors and brings up a failure message although the electronic parts in the detectors are working just fine. The team now added a stiff strand to the sistem that allowed to fix the distance of the detectors on deck in a way that their distance will not be altered when pushed down the stem tube - not even a tenth of a millimeter. This made the sistem work.

Additionally it was found that the sistem is incredible sensitive not only for distance alterations of the sensors but also to resistance alterations on the contacts. Making 1 tank work can require the pulling out and reassembling with tenth of millimeter precision and exact pressure on the contacts a dozend of times until the sistem recieves a signal from the stem that is accepted as "valid signal" from the electronics board.

Imagine the communication between the detection stem and the circuit board rather like a computer detecting a "peripherical device with exact characteristics" instead of recieving a simple signal. Dedectors can be fine, cable can be fine, and still the signal can be rejected as "present but not valid".

In the case of the 2 repaired tanks we where able to restore the stem with original detectors that where on board to the point where the signal sent by the stem was accepted as valid.

With the 2 remaining tanks we ran out of supply of original detectors - we where thinking that by adding the distance holder strand - we got to the core of the solution for a stem that sends acceptable signals - and we can fabricate detectors just as Ineleth does, from market avaiable reed switches and resistances and then assemble those detectors to stems that send acceptable signals to the circuit board improving the original complicated way of assembly by a "distance holder strand" that would allow easier assembly.

This was the status when we had the first call to Klaus at 3.00 PM LT. We had successfully "inverse engineered" the whole sistem had reasembled it to "signal acceptance" in two occacions successfully, using surplus reed switch resistance resin sensor assemblies fabricated by Ineleth and thought that we have the insight to rebuild those sensors from standard parts from cero just as the original manufacturer and then assemble them to a acceptable signal stem.

It turned out that the manufacturer seems to have built in a kind of "additional glitch" that avoids doing exactly that (probably more for marketing than for security reasons) You can not present a "built sensor" to be accepted as "original peripherical" to the circuit board. We built 23 versions of the sensor and could not get it accepted as "original peripherical device". What we promised for "noon the next day" fabrication of a functional stem to be accepted as "original and porperly working" by the circuit board, using standard market electronic parts, could not be fullfilled. I called Klaus when being 3 hours behind the promised shedule admiting defeat in getting the left 2 tanks working without "original manufarturer deteccor cast assemblies necessary to build functional stem components".

Outcome and recommendation to get rid of the sensor failure problem once and forever and make it an "easy fix" for the crew.



- Your chief has a deep understanding is a competent man and obviously has a long history of dealing with the problem.

- What is written here, he already knows it, and he is on top of the problem with ordering "complete sensor stems" to Savannah from
Ineleth what will solve the problem without trouble according to the understanding of all involed.

- There is no chance for an average electrician to deal with the complexity of the sistem when obligated to break it down to to the level of sensor change, only stem change can be performed.

- The sistem is a design "calling for trouble" so if you can insist in design improvements from Ineleth please do it - and if you can not get it - keep lots of pre-assembled spare stems available at all time - you will need them.

Water in the sistem, as it was detected at opening, will ruin the disconnect sensor wires assembly and alterate it to the point of "non signal acceptance". Then the water will run straight down the stem tube and mess with the extremly sensitive sensors and contacts there. Movements of cables and sensors inside the stem, - even the most subtile - will make the sistem fail - so will oxidation. A better design would be to cast the 2 sensors and the cables in resin and work with a "stem inside the stem approach" what we did with the distance holder cord. A filling of the whole stem tube with resin would also be good, avoiding both movements and water running down, which can both bring up, non signal acceptance. When replacing the stem tube, embedding the "disconnect sensor wire assembly" especially its contacts in loctite, epoxi or similar is recommended.

Even if the stem turns out a proper peripherical signal - it is a long way to the circuit board on the bridge and any alteration of the signal by oxidation of contacts or whatever reason will make the sistem fail which is specificly designed for a "extreme narrow signal acceptance window". Any attempts to manipulate the sistem below the level of a "complete detection stem replacement" are not recommended.

 

OUTCOME

When we came on board the following tanks where failing

1p
5p
5s
slop stbd

The outcome of the effort was that

1p
5p
where deletd by GL - as we repaired it with the originals sensors (without alterating the crappy resin less base design of the stems) to the "original status with orignial parts".
5s
slop stbd
should be a doable fix for your very competent chief in Savannah with pre calibrated stems from Ineleth - also going for complete acceptance by GL.

As long as the "crappy base design" is not improved by Inelteh, future failure will occur on all tanks, be prepared with lots of spare stems and your Chief informing others how to get rid of the problem.

 

 



__________________
Page 1 of 1  sorted by
 
Quick Reply

Please log in to post quick replies.



Create your own FREE Forum
Report Abuse
Powered by ActiveBoard