Railcar 9 arrives on site. Sample drawn from car measures 14.6 m. Put another sample into tester and leave over night.
Transfer last trailer from railcar 8 to detector.
Transfer two trailers from railcar to detector. Empty railcar into thrid trailer and release railcar. Store trailer in Site 40 over weekend.
New flexible hose installed. Transfer one trailer from railcar to detector.
Discover flexible hose from railcar has leak in it. Stop transfer. Order new hose.
Transfer lsat trailer from railcar 7 to detector. Move plumbing at railhead. Take sample of railcar 8 for testing.
Finish unloading railcar 7 and transfer 2 trailers to detector. Last trailer sent to Site 40 for the weekend.
Railcars 7 and 8 arrive. Test sample from first car and it measures 23 m. Transfer one trailer to detector.
Accept railcar GATX89825 and reject railcar NATX230206. Ask Boncosky to hold shipment until we have started offloading railcars 7 and 8.
Retest NATX230206 sample. Attenuation length is 11.0 m. Jen completes Cincinnati tester test of batch sample. It looks normal.
Measure second sample from railcar NATX230206. Attenuation length is 11.4 m. Fill with and test batch sample. Attenuation length is 15.5 m. Refill with NATX230206 sample and leave over weekend.
Receive second railcar sample. Test sample from NATX230206. Attenuation length meaures 10.8 m. Fill with sample from GATX89825. Attenuation length is 16.5 m. Refill with sample from first railcar and leave over night.
We receive one two gallon railcar sample from NATX230206 and the batch sample. Put the raqilcar sample in the tester and leave overnight.Railcars 7 and 8 finally released from Houston.
Jen and Randy visit Penreco in Houston and observe NATX230206 being filled. Detailed summary of visit to follow. Two cars filled and samples shipped.
Give go ahead to release railcars 7 and 8 from Houston.
Measure batch sample. Attenuation length is 15.5 m. Refill with sample D2005B. Measure attenuation length to be 16.4 m. Measured a second railcar 6 sample and got an attenuation length of 24.4 m. Start microbubbler tests.
Labels on sample are confusing. Sample in tester is sample D2005A. Attenuation length is 18.3 m. Refill with sample D2005B. Attenuation length is 15.0 m. Refill with batch sample.
Jim Morris runs out of time and does not visit Fermilab. Oil samples from Houston arrive. We put one in the alabama tester and leave it over the weekend.
The next two cars were filled today. Samples sent to FNAL. We arrange to have Jim Morris visit Fermilab.
We had a conference call today with Houston. We agreed on the following:We volunteered to go to Houston to observe the next filling of the railcars. Later it was found out that the filling would occur on Thrusday and we could not get there on time. We are invited for the next filling however.
- We would take a day longer to test the samples allowing the samples to settle before testing.
- We would receive two gallon samples from each railcar.
- The samples would be extracted from the railcars after the filling process was complete (and possibly after a little settling occurred in Texas) rather than during the filling process as is the present practice.
- Houston would try to spray more of the car with mineral oil when they initially spray a little in a drain it out. (This actually couldn't be done.)
- Houston would drain more oil from the bottom of the car before filling our saample jugs.
Final trailer from railcar 6 is transferred to detector. Railcar 6 released.
Two trailers are transferred to detector.
Railcar 6 arrives at Fermilab. Attenuation length is 22 m. First trailer is filled and transferred to detector.
Waiting for railcar 6.
Railcar 6 is supposed to arrive at FNAL today. Remeasure sample from UTLX663312. It measures 11.1+/- .1 m. Batch sample measures 16.3 m which is about standard for a batch sample. Conclude that samples in the cars are contaminated. Reject this load.
Measure attenuation length of railcar samples. Railcar UTLX663312 measures at 9.9 m and railcar UTLX663313 measures at 8.3 m. Awful. We will remeasure these samples tomorrow.
Samples arrive late in day.
Two cars are filled in Houston; samples sent.
The final trailer from railcar 5 was filled and transferred to the detector. Railcar 5 was released.
Transferring continued today from railcar 5. Two trailers were loaded and were transferred into the detector.
Railcar 5 has arrived; railcars 3 and 4 were picked up. Two 55-gallon drums were filled from the car and two samples were taken. The first sample measured 22.3 m. One trailer was filled and transferred to the detector.Houston released the railcar that had a reasonable attenuation length.
Eric examples the sample from D2001C and notices that water has accumulated in the bottom of the sample. Jen also put that sample into the Cincinnati test and see water in there also. (We are going to have trouble getting that out...See Friday's problems with water in the trailer.) We accept the railcar corresponding to D2001D and reject D2001C.The final trailer from railcar 4 was pumped into the detector. According to the pressure on the bubblers, we have 16.6 ft. in the detector. This corresponds quite well with what we think we put in. We think that we have 90750 gallons in the detector 750 gallons in the overflow tank, 935 gallons in drums and about 50 gallons waste.
No sign of the next railcar. Railcars 3 and 4 are still at the railhead.
Jen measures the samples from the latest batch. The batch sample in the Cincinnati tester looks normal. The first of the railcar samples, D2001D, has an attenuation length of 18.0 m. The second sample, D2001C, looks very cloudy when put into the tester. After sitting for a while, it measures an attenuation length of under 1 m. On retesting D2001D, the length came out to 18.4 m.
The trailer with the water in it was tipped and the water adhered to the bottom of the trailer. That trailer was taken to Boone and the oil was transferred to the detector. Again the water adhered to the side of the trailer. It was then taken back to Site 40 and swabbed out. In the mean time, the second trailer was filled at the railhead. It was transferred to the detector after the first trailer. Finally, after the swabbing, the first trailer was filled at the railhead draining railcar 4. That trailer was taken to Site 40 to sit over the weekend.At the end of the day, we received the next set of samples. The first was put into the Alabama tester.
The trailer with the snow in it was taken to Site 40 this morning and heated up. The trailer that was in Site 40 was brought over to the detector and transferred in.
A final measurement of the attenuation length of a railcar 4 sample (jug 2) was made this morning. It measured 22.0 m. With that, we decide to start the railcar unloading. Two trailers were filled and one started to be unloaded when the snows came. After about 4000 gallons unloading, Eric noticed that snow was blowing into the top of the trailer and suspended transferring. (A small amount of water was visible at the bottom of the trailer.)
The pumping equipment had to be moved up-track to reach the railcar 4. After that, samples were extracted (they looked clear and free of water). Two 55 gallon drums were filled, and two gallon jugs were filled. When the first sample was put into the attenuation tester, it looked very hazy at the top. When this had settled out and the oil looked clear, we measured the attenuation length. It measured 12.3 m!! A sample from the second jug was put in the tester. It measured 21.4 m. A second sample from the first jug was put in again and now it mesured 17.4 m. Finally a sample was put in to sit overnight. It will be measured tomorrow.
About 10 gallons was extracted from the lower line of the detector today. In the early part of the extraction, black sandy particles with a little water came out. By the end, the oil was basically clean. Oil was also extracted from the trailer in Site 40 and that too seemed to be free of water. With that information, we decided to go ahead with the oil fill. The last trailer from railcar 3 was loaded into the detector.Eric and Jen found that the lens in the Alabama tester was bumped out of alignment. This may have accounted for the high curvature results that Randy saw on Saturday. The railcar 2 sample was remeasured and it had an attenuation length of 19.2 m and a normal shape. Given this, we conculde that the alabama tester is back on-line.
The oil that we put in the Alabama tester from railcar 3 Friday indeed had water droplets in it. We could also see them on the bottom of the tester and in the jug. A quick guesstimate of the amount of water we transferred from the railcar given the amount of water in jug 3 (before we started transferring) and jug 4 (after the railcar was almost drained) and given the approximate size and number of the drops is 25 cc's. That could be off by as much as an order of magnitude, but I'm pretty sure we have less than a liter of water in the detector. The water should mostly settle to the bottom over the weekend. We will back pump some on Monday and see if we can remove it.I cleaned the water out of the Alabama tester and measured the oil from railcar 3 again. It had a very long attenuation length (33 m). This measurement was a little bit suspect in that there was an abnormal amount of curvature to the plot of light vs. length. I also measured a second sample from railcar 2 and it also had a long attenuation length (30 m) and the high curvature. I would like Eric and/or Jen to check the Alabama tester on Monday to see if I screwed it up when I cleaned it.
Today we emptied railcar 3 into three trailer (one of the two was loaded twice). Two of the trialers were transferred to the detector. The third trailer was sent to site 40 to sit over the weekend.Jen remeasured the sample taken from railcar 3. It still had a very long attenuation length of 23.7 m (previously it measured 20.6 m). Another sample was taken toward the end of emptying railcar 3. When we put thaqt sample into the Alabama tester, we again noticed small water droplets on the bottom. After that, we noticed them in the bottle also. This means that the oil from railcar 3 probably still had small water droplets in it for the entire portion of the fill!!! The water should collect in the fill line. We will institute a pumpout procedure of this line every so often to try to remove this water. We will have to do a better job of inspecting the oil in the future.
After consulting with Boncosky, the railcar in Houston with the better of the attenuation lengths was released (it will start on its merry way on Monday). Exxon will give us a proposal about what ot do with the other car. Options are: (1) they will sell it to someone else, (2) they will re-refine it into something else, or (3) they will send us another sample from it to test and hope that the sitting magic works on it.
After allowing the oil in railcar 3 to settle over night and then extracting approximately 20 more gallons, a sample was taken that seemed to be water free. A trailer was filled and transferred into the detector. We could have filled and transferred another but the drivers had a "union" meeting.With regards to the oil samples from batch 3, Jen measured the GATX railcar sample again (with fresh oil). It improved a bit to 14.2 m. However, this change is consistent with the previous measurement of 12.3 m. She also measured the batch sample taken out of the Cincinnat tester. It had an attenuation length of 14.5 m. Finally, Jen remeasured the sample from railcar UTLX 663338 and it was 18.7 m (it measured 21.7 m the first time -- again consistent).
We discussed batch 3's numbers with Boncosky. We are holding up the decision as whether to accept or reject this batch for another day. We will explore options tomorrow.
Overnight two railcars were delivered to Fermilab. In the morning, first two 55 gallon drums and then a sample was extracted from the first (dubbed railcar 3) railcar. When that was put into the tester, it looked extremely turbulent. After it settled down, it was noted that there were water drops on the bottom of the tester. After re-examining the sample jugs, water was found at the bottom of these also (remember this was after ~100 gallons was extracted from the car). The tester was cleaned, another 20 gallons was extracted into a 55 gallon drum, and another gallon sample was taken. This may have had a few droplets of water at the bottom. Eric carefully decanted enough oil to fill the Alabama tester. Later, Jen measured the attenuation length and it turned out to be 20.6 m (this was from the samples which were about 15 m when we received samples from Houston).Testing also continued on the three samples received from Houston yesterday. The batch sample was tested in the Cincinnati tester and it looked normal. The other railcar (UTLX 663338) was tested in the morning before the problems with railcar 3 and it had an very good attenuation length of 21.7 m.
Samples from shipment 3 were received. One railcar sample (car GATX 14424) was measured and had an attenuation length of 12.3 m.
Two more cars were filled in Houston. Samples were Fed-Exed to Fermilab.
We had a telephone conference with Houston today. The cars had not yet left Houston, but the necessary paperwork was was sent such that they would be on their way. The next shipment will be filled on Monday, January 21 and samples will be Fed Exed to Fermilab.
No signs of oil leaks were seen on the manhole flange. The oil cars were released.Oil delivery operation shuts down until next cars or samples arrive.
The final trailer from railcar 2 was transferred into the detector today. Thus all of the delivered oil to date is in the detector. The manhole was re-examined today. There may be a thin film of oil where the drop appeared yesterday. Techs check the torque on the bolts and everything seems fine. We decide to check again on Monday, and, if everything is OK, we will release the railcars at that time.
Two trailers were transferred into the detector today and the final oil from railcar 2 was pumped into a trailer. A small drop of oil was found on the manhole leading to worries that the manhole might be leaking. The drop was very close to the area on the gasket where the rope was doubled over. The flange was cleaned. We will wait and examine it again tomorrow. Even though the railcars are empty, we decided not to release them today because we are afraid that we might have to replace the gasket again.
The first trailer from railcar 2 was transferred to the detector today. Two trailers were filled from railcar 2.The filtered sample from railcar A was tested today. The attenuation length went up to 13.9 m. The unfiltered sample from railcar B was retested. This time it had an attenuation length of 17.5 m
There is no signs of leaking from the new gasket. The two trailer loads of oil in Site 40 were pumped into the detector. The pumps were moved from railcar 1 to railcar 2. The first empty trailer was filled and moved to Site 40 for heating.The second railcar sample (railcar B) and the batch sample was tested in the Alabama tester today. The former had an attenuation length of 15.2 m and the later, 13.0 m. We called Boncosky and delayed shipment of the oil. We had Anna filter the sample from railcar A and will test it tomorrow. The batch sample was tested in the Cincinnati tester. It looks normal.
The manhole flange was retorqued. One trailer was slowly pumped into the detector. On overtime, that trailer was refilled with the oil in railcar 1 and taken to Site 40 for heating. Railcar 1 is again empty.The samples from Houston arrived today. One of the railcar (railcar A) samples was put in the Alabama tester and measured in the evening. It had an attenuation length of 13.6 m. A sample from the other railcar was put into the tester, will be allowed to sit overnight, and then tested in the morning. The batch sample was placed in the Cincinnati tester and also allowed to sit overnight.
The new gasket was installed today and the manhole sealed. Monday, it will be retorqued and then we will start filling again.
The rest of the oil was pumped out of the detector today; one more trailer load was taken to Site 40 and the rest was pumped into the overflow tank. We approximately half-filled the overflow tank.In the afternoon, the manhole was opened. It seemed that the only place the gasket failed was near the one bolt. It may have had to do with some pitting in the flange at that point.
Approximately 500 gallons was pumped back into the detector and the veto shield plate was re-installed.
In Texas, two more railcars were filled with oil and samples were sent to us.
In response to the oil leak discovered Wednesday, we began removing the oil from the detector today. The direction of the pump cart at the railhead was reversed. The two trailers that were filled from railcar 2 were emptied back into that railcar. We then started pumping oil out of the detector. Two trailer loads were pumped out of the detector, one of which was pumped back into railcar 1 and the other is stored in Site 40. We presently have about 9000 gallon left in the detector. The plan is to fill the other trailer with 7000 gallons tomorrow and pump the other 2000 gallons into the overflow tank. We can then open the detector again.On the new gasket front, Peter K. found out that material was not available for the flat gasket and decided to use the Gore rope gasket instead. Fermilab has used this type of gasket on other application and has been pleased with the results.
The gasket material should be here by tomorrow. The expectation is to open the detector tomorrow afternoon, inspect the old gasket, refill with about 600 gallons from the overflow tank, and then seal the detector Saturday morning. Monday will start the refilling process in earnest.
The next oil shipment has been delayed. The people at Pennreco failed to get the next railcars to the yard yesterday. When they got them in this morning, they found that the cars had water in the bottom and had a musty smell. They are to be swapped for two new cars tomorrow. Samples will either be drawn Friday or Monday and we will test them the following two days. The cars will then be released next Wednesday or Thursday with delivery late the following week.
HAPPY NEW YEARToday, disaster struck. The seal at the manhole was discovered to leak. We had filled the detector with the last trailer load of railcar 1, had filled a trailer from railcar 2 and had started filling the second trailer from railcar 2 when we decided to halt operations. We will be reconstructing the pumpcart at the railhead tomorrow such that we can pump into the railcars. When that is complete, we will pump out the detector into the railcars. We expect to have the detector empty by the weekend. Wish us luck.
Jen tested the sample taken from railcar 2 on Friday and it had an attenuation length of 24.2 m, a great New Year's present for us. Looks like we are ready to go full bore on unloading railcar 2!
A trailer was emptied into the detector and the other trailer was filled at the railhead. This emptied the first railcar, a real milestone! We have now one trailer sitting over the long weekend in Site 40. Len did not warm this car; we will see how warm it becomes just sitting in the building. The pump trailer was moved to the second railcar and two 55 gallon drum samples were extracted followed by a testing sample.
The first major transfer to the detector occurred today. First, approximately 1800 gallons from trailer 1 was transferred at a slow rate (~10 gal./min. for the first 1000 gallons and then 15 gal./min. for the remainder). We had espected that this truck carried 2500 gallons and the pump ran dry when the trailer emptied before we expected it to.The seal was inspected while the trucks were being shuttled. It was dry.
The second trailer was emptied in the afternoon and early evening. By the end of the day, approximately 9300 gallons of oil were in the detector.
While trailer 2 was being emptied into the detector, trailer 1 was being reloaded and heated. It will be transferred to the detector tomorrow.
The manhole was sealed this morning. Trailer 1 was filled with 2000 more gallons (making a total of 2500 gallons in it) and warmed in Site 40. The good news on this is that the oil still flowed from both the pump trailer and the trailer even though the temperature got down to about 14o F over the weekend. Trailer 2, which sat in Site 40 over the Chritmas holiday, is now at about 64o F.
Today we pumped out 150 gallons from the detector to make sure that the emergency pump worked. After finding the correct circuit breaker, it did. The emergency pump had a small problem in that it leaks a little bit around the impeller shaft. We will decide later whether or not to replace it immediately, wait to swap it out for the one in the pump cart at the end of the fill or live with it the way it is.The disk completing the outer wall of the veto is in place. Everything is shut down for the long weekend.
NEWS FLASH --- OIL IS IN THE DETECTORThe big event today was the moving the oil from the overflow tank into the detector. It was not a problem and was actually kind of fun to watch the oil come in. The index of refraction of the oil first seemed to make the base of the flow pipe rise as the oil level rose. Then things started disappearing. By the time we got the full 600 gallons in (that came up to 11.5 in from the port) we could not see the PMTs at the base.
We took samples from the bottom of the fill pipe and siphoned some from the detector. Now begins a strange saga. As you remember, the samples we took from the overflow tank all measured about 8 m attenuation length on the first day. Yesterday we took another sample from the overflow tank. It measured 13.2 m. The sample we took from the bottom of the stand pipe today measured 13.3 m. And finally, the sample we took from the detector measured 20.9 m!!!! While I can't explain it, it looks like we have good oil in the detector!!!
We have filled trailer 2 completely. It is now over in site 40 being heated. The oil in the railcar had gotten down to 37o F at the bottom. The oil in the middle of the trailer was 47o F. By heating it with 110o F water for about 6 hours we brought that up to 57o F. Tomorrow, we will continue to heat it to about 65o F.
The next oil delivery is delayed. The railcars will be filled on Jan. 2, we will receive oil samples on Jan. 3, and OK the shipment on Jan 4. They will then roll from Texas on Monday, Jan 7 arriving at FNAL Jan. 15-18.
Today we degassed the overflow tank sample (it measured about 9 m also), and took four gallon jugs from the overflow tank for long term tests. The oil has now been in the overflow tank since Friday. The sample we extracted today measured 14 m up from the 7.5 m measured on Monday.We ran the sample from trailer 2 that was taken during the transfer to the overflow tank through the Cincinnati tester (it looks good) and compared it with a sample taken from the overflow tank. Aside from an overall normalization shift, there only seems to be a little depression at 320 nm -- right where you would expect it from dissolved air. We do not seemed to have done anything chemically to the oil.
With that conclusion, we decided to bring oil next to the detector. This evening we filled the overflow pipe and are letting it stand overnight. Tomorrow, we will take a sample from the bottom and if it seems to be good, go ahead and put oil in the detector.
The pumping system has worked well both today when we filled the stand pipe, when we took samples this morning, and yesterday when Rex and Co. were calibrating meters and taking samples.
Not much work has progressed on the heating front. The electricians that had to put a new line into Site 40 had an office party today. They will do that tomorrow. We will then fill trailer 2 and run a heating study on it.
The news on filtering is not good. The vendor who was going to rent us a unit has come back with a unit that does not look like it will meet our needs. A second vendor is willing to sell us a unit with polypropylene filters...a filter I don't think we want. The third vendor has a spun glass filter that would be perfect, but only has anodized aluminium housing, cannot get us the unit until Jan. 8, and wants $11K for it, effectively taking them out of the picture in my mind. Luckily, I worry less about crud now that we have seen good attenuation lengths in the trailers.
Today we measured the oil that was taken from the overflow tank and degassed. It too measured in the 8 m range. Quite discouraging. We then took the sample from trailer 1 that had measured 21.6 m on Moday and shook it violently. After letting the visible air bubbles settle out of it, we remeasured it and it had a 9 m attenuation length.Jen was about to measure a new sample from Exxon (the prelude of a new shipment) when disaster struck. She slipped, spilled a good deal of oil, and cracked an epoxy joint on the test tube of the Cincinnati tester. She was able to fix the tester, but was not able to do as many tests with it as she would have liked before she had to leave for break. She did test the new oil (looks pretty much like the present oil).
Today we investigated filtering systems for the oil. We have received a bid from a Texas firm which made some particulate matter measurements on the oil over the weekend. They see relatively high amounts of particulate matter in the oil but essentially no water. They advise a filter system. We have also contacted a local vendor and will speak with them tomorrow. The funny thing about these measurements are that we sent them an 8 m, a 16 m, and a 24 m oil sample and the particulate counts come back almost the same. Anna Pla's measurements are very similar.On Friday we extracted a sample from tanker 2 midway in the fill of the overflow tank. Over the weekend, Jen measured this oil in the Cincinnati tester. It looks very similar to the oil that we have seen before from Exxon. Eric also measured this oil in the Alabama tester and it had an attenuation length of 24.1 m.
We extracted two 55 gallon drums of oil from tanker 1 and then drew 4 gallon samples. The attenuation length in one of these samples was 21.6 m (up from the cononical 9 m we were measuring before!).
Finally we measured two samples from the overflow tank. The first measured 7.5 m. We then asked Anna to filter the sample for us. It then measured 6.8 m, not a significant difference.
Tomorrow, we will try degassing the overflow tank sample. Other than that, I'm not sure what to do.
Also on the very near horizon, we have to decide whether or not to delay the next shipment of oil. Since we are not pumping now, I don't see how we can be be ready to take another shipment on January 7.
OIL WAS TRANSFERRED FROM TRAILER 2 TO THE OVERFLOW TANK TODAY!!!We first moved trailer 1 to the railhead and brought trailer 2 back to the detector. We flowed 100o F water through it for about 1.5 hours bringing the temperature up to a comfortable 64o F. We then took a short sample from the trailer to inspect visually (it looked OK), filled a 5 gallon container to prime the pump, and then hitched the trailer to the fill line. After priming the pump and having it run for about 10 minutes (and just a little after Camilo expected oil to flow) it started to run. The pump was able to pump at a rate of about 35 gallons a minute. We took a two jug sample from the trailer after about 1/2 of the volume had entered the overflow tank. We then run the pump until the trailer ran dry (the techs at the trailer could hear the pipe sucking air). After closing off the trailer, we tried pumping oil from the overflow tank back into itself. We had a bit of a problem priming the pumps for thsi direction, but finally it work. Camilo was thanked and he headed for his flight back home. We also took a two gallon sample from the overflow tank during the pumping process which we will test shortly.
On the attenaution front, we sent a sample to a filtering firm in Texas. They plan to test the sample for water and particulate matter over the weekend and tell us what type of filtering they would recommend on Monday.
It looks like we will be able to heat the trailers in Site 40 overnight.
NEWS FLASH
We just measured a 19.9 m attenuation length from a sample that was taken from trailer 2 and degassed. We just may be seeing the effects of dissolved air in our trailer samples (the difference from 15 m to 20 m). That also may be the reason the filtered sample from trailer 1 only got back to 15 m. Things may be making a little sense.
Today we tested four samples and are about as confused as ever. There seems to be very little rhyme or reason to the attenuation length measurements. First to summarize today's measurements:
- We drained trailer 2 and half way through the draining process took a sample. Results: attenuation length = 16.4 m (yesterday's initial measurement was 14.6 m -- consistent).
- Later in the day, we filled trailer 2 with 700 gallons and drove it around. We then took another sample. Results = 15.8 m; again consistent.
- Earlier in the day, we put part of the sample from yesterday's pump trailer sample (attenaution length = 8.9 m) under vacuum. Lots of foam formed on the sample (the pressure in the chamber went from 200 to 300 microns to 45 microns). However, the attenuation length got slightly worse (6.9 m). We also tried to outgas the remainder of railcar 1 jug 4 (attenaution length of 26.6 m). It showed very little foam, started at only 75 microns and went to 48 microns after 45 minutes. We did not retest this sample.
- We took an additional sample from the pump house immediately after filling trailer 2 for the second time. This sample had an attenuation length of 5.4 m.
That is the data; the conclusions are hard to draw. The three samples of pump house oil have all measured very low (8.9 m, 6.9 m, and 5.4 m) event though the oil that was extracted from the trailers the pump house filled mesured better (in particular, the 5.4 m measurement should have been oil delivered to trailer 2 which measured subsequently at 15.8 m).
The three samples we have measured from trailer 2 are consistent at 14.6 m, 16.4 m , and 15.8 m. This is significantly above the oil from the pump house either immediately before the trailer fill or immediately after it.
The first trailer samples were consistently at a much shorter attenuation length: 6.1 m, 9.1 m, 7.8 m, and 9.3 m. These numbers improved to 14.9 m after filtering.
I'm baffled. Any ideas?
Today we tried to track down the cause of the poor attenuation length in the trailer1 sample. In doing so, we ran attenuation tests on seven oils, a new record for the daily output for the Alabama tester. Jen and Eric deserve a round of applause. While we have a few clues, we do not have a smoking gun, and we can only guess at what is our problem.After yesterday's report, Jen tested a sample we took from trailer 1 in the evening. That too tested in the 8 m range. She then left a second sample from that jug in the tester overnight and tested it this morning. It tested at about 9 m.
Hoping to verify our test procedure, we took a third jug directly from the railcar (while bleeding only about a gallon into the pail before sampling) and tested it. It had the normal curvature, but measured at only 12 m instead of the 20 m we had measured on Monday. More on this later.
Next we tested a sample from trailer 1 (which measured about 8 m yesterday) that Anna Pla had filtered for us. It tested to be 15 m, a definite improvement. Anna had some problems with the filtering (I have no details) and said that she would do it again if the oil still had problems. We have not taken her up on this offer yet.
We then went back to a sample we had retained from Monday's extraction from the railcar (which measured 20 m). This new sample measured 25 m. Eric hypothesized that the jug we extracted earlier in the day from the railcar might have only sampled oil left in the valve, plumbing, and very bottom of the car and that is what gave the short attenuation length. With this in mine we drew another 55 gallon drum of oil from railcar 1 and then filled two more jugs (4 and 5). Testing jug 4 gave a whooping big 26.6 m attenuation length.
Earlier this morning, we put about 50 gallons into trailer 2 and drove it around the lab site for about 1/2 hour. I then extracted a gallon from this. This gallon, unfortunately was probably only from the plumbing. However, it measured 15 m.
The final measurement for the day will be a sample that has sat in the pump cart hose overnight. We'll have those numbers tomorrow.
We are about at the end of what we can test. More ideas are welcome. The data may be consistent with particulate matter in the oil that slowly settles to the bottom. In this case, we should filter it before putting it into the detector. A better volume to surface area ratio in the trucks may also help. We may have been hurting ourselves by transporting such small volumes. Tomorrow we will have to decide how to proceed.
On the oil heating front, Jesse has found a 10 kW heating system on site. We had it working on trailer 1. We have also ordered a 18 kW heater. Tomorrow we will check with Exxon as to how high we can heat the oil right around the heating coils. That will determine our energy transfer rate. FNAL will put concrete blocks into lab E/F to store the trailers indoors while they warm up.
This morning we began extracting a small sample from the railcar and visually examining it. It looked clean. Then after a few fits and starts with the railhead pump, we transferred 700 gallons into the first trailer and headed to the detector.At the detector, we extracted and discarded two two gallon samples: the first of which had a little grit in the bottom of the bucket, but the second one looked pretty clean. We then tried to extract a 55 gallon drum sample, but the oil was flowing so slowly that we only extracted about 15 gallons into the drum. We then took 2 one gallon samples for testing. When we disconnected the sample hose from the trailer, a little grit ended up in the bucket.
The testing did not go as well as yesterday. The first sample was introduced into the Alabama tester and allowed to sit (and warm up) for over 1.5 hours. The HV was turned on and the tube was allowed to equilbrate for about a half an hour. The attenuation length measured in this test was 6.1 m with severe negative curvature (remember in most of our tests, we see a small positive curvature). The 1 meter attenuation length was about 4 m and the 0 meter attenuation length was about 11 m. At this point we terminated transfer operations for the day and set about remeasuring the oil. The second sample was better, but only relatively. It measured as a 9 m attenuation with less negative curvature (12 m at 0 m and 8 m at 1 m).
We are puzzled by these results. Eric sees small flecks of dirt in the samples that we have tested. To follow this up, we are asking Anna to filter the second gallon jug that we extracted this morning. We also extracted a third gallon jug from the trailer this evening and Jen will measure it later tonight. If the attenuation length is still poor on the last sample that we extracted, tomorrow we will try filling the other truck with 700 gallons and seeing if that one is any better.
Vsv. heating the oil: We are converging on a system that will heat the oil in the trailers. We would fill two trailers in the afternoon, take them to Site 40 and park them inside overnight. We would circulate hot water through the trailers (with the heat exchanger that Jesse found, we could probably raise the oil temperature by about a degree an hour). We would then transfer the warmed oil.
A little after lunch today, we received permission to start the oil transfer. Mucho Gracias to Eric, Jesse, and Camilo for getting this through. At 3:00 we had a tech meeting to go over the transfer procedure and have each person sign off. If you are to be involved in the oil transfer, you have to see Eric, read the job hazard report and the transfer procedure and sign Eric's list. We can get in real trouble if you play with any of the oil plumbing valves and you are not authorized. Play smart for the experiment and get authorized.At about 4:15 we started the first steps in the transfer procedure. We hitched up the plumbing to the tanker car and took some test samples in a bucket. The first two gallons had about a quarter cup of water in it and a lot of dirty crud (suspended dirt particles). We discarded that sample, and took a second sample. That sample had a few drops of water and a lot less crud. After discarding that sample, we took a third 2 gallon sample. That one had no visible water in it and only two or three pieces of suspended dirt. We discarded that sample also and filled two 55 gallon drums. We then took samples in two gallon jugs for attenuation testing.
Len measured the oil temperature in the 55 gallon drums and the one gallon jug. In both cases, the oil temperature came out to be 8o C.
We measured the attenuation lengths in the Alabama tester. The two runs came out 18.6+/- .4 and 21.7 +/- .5 giving an average of 20.1 m.
The plan for tomorrow is to fill the first truck with 650 gallons of oil, test it, transfer it to the overflow tank and let it warm up a bit. We will either put the first oil in the detector late Tuesday or early Wednesday morning.
Tanker cars moved to end of track.
First two railcars arrive in early morning. Oil temperature measured at 64.5o F from top of car with thermocouple. Cars not positioned correctly on track. Request that they be moved.