ASME B31.3 vs ASME B31:4: Usage and Calculations For Ore Concentrate (Slurry) Pipeline 2

[jackieray]

Dear all,

Greetings everyone. I have something I want to ask about pipeline construction code.

I have recently started working at a mineral processing company that produces ore concentrate. The concentrate slurry (water and ore mixture, 66% solid) is transported over 62 miles (100 km) through the slurry pipeline to a concentrate dewatering facility. I am now tasked with a project to determine the cause of accelerated wear to the pipeline which, according to the senior engineers here, cuts the service life of the slurry pipeline to almost half than what it used to be (15 years decreased to 7 years).

Before tackling the main bulk of the project, I discussed about the current method of wear calculation with the pipeline engineers. They said NDT is used to measure pipeline wall thickness wear for each mile (mile 1, mile 2, mile 4, ..., mile 62). However, due to the mountainous terrain of the pipeline location, the NDT measurement that was started only a little over 1 year ago cannot be done regularly (IMO, even quarterly NDT measurement for each mile is near impossible). Also, some part of the pipeline is buried underground, adding to the difficulty of the NDT.

The engineers have calculated the minimum allowable thickness (MAT) for each mile since there are differences in concentrate pressure (specified by pipeline constructor). Here are some questions I want to ask to the distinguished engineers here.

1) I noticed that they are using ASME B31.3 to calculate the MAT. As I have read previously on the threads here, the B31.3 is said to be used for process piping. Should the MAT calculation be redone using ASME B31.4 which is specified for pipeline transportation for liquids and slurries?

B31.3 MAT formula : t = PD/2(SEW+PY)
B31.4 MAT formula : t = PD/2S

2) I have tried calculating the MAT using both codes. I found that the MAT determined using B31.3 is well thicker than those of B31.4, which is in line with what the members of EngTips found. However, I also found that my calculation for MAT using B31.4 is also thicker when compared to the calculations of the pipeline engineers, which used ASME B31.3 as their code.

Can anyone point out my mistake in calculating the MAT value? Below is the data of the pipeline that I used (values determined using B31.4 unless otherwise noted).

- Carbon Steel API 5L X60, seamless, NPS 6", Sch 80, D 6.625 in, d 0.438 in (by most standards it should be 0.432 in but 0.438 is what the engineers here use)
- Slurry operating temperature (T) = maximum (approx.) 100 F (operational data)
- Quality factor for seamless pipe (E) = 1
- Weld joint strength factor for T<400 F (W) = 1
- Coefficient for ferritic steel & T<482 F (Y) = 0.4
- Specified minimum yield strength for API 5L X60 (Sy) = 60200 psi (from API 5L Specification)
- Design factor (F) = 0.72
- Basic allowable stress (S) = approx. 43300 psi (calculated using S = S x E x Sy from ASME B31.4)
- Pipeline measured internal slurry pressure (P) = up to 4500 psi (on steep hills, value determined by pipeline conctructor)

Using the B31.4 formula for MAT, I found that the MAT is around 0.344 in (8.73 mm) for 4500 psi of slurry pressure.
However, the pipeline engineers' calculation using B31.3 formula for MAT is 22% lower, around 0.270 in (6.87 mm).

Thank you in advance, everyone.

 

[DekDee]

jackieray,
Only post in one forum please - eng-tips rules.
Cheers

 

[1503-44]

B31.4 is for hydrocarbon liquid pipelines. Slurry pipelines should be designed using other methods. possibly B31.3, at the owner's option, but noting that your pipeline is not a process.

S for B31.3 is not 60,000 psi. 60,000 psi is a yield strength, not an allowable stress.
You will find S for 31.3 in the pipe material table in the B31.3 code.

As for your erosion problem, uour slurry is very abbrasive, or your velocity is too high, or both.


--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[jackieray]

@DekDee

I'm sorry for the double post, and thank you for the warning. I have deleted the other thread

@1503-44

1) Is B31.4 not suitable for ore slurry?

I do understand that B31.4 used to be named "Pipeline Transportation Systems for Liquid Hydrocarbons and Other Liquids" but since it's now (2022 version) termed "Pipeline Transportation Systems for Liquids and Slurries", I thought that it can be used for ore slurry pipeline systems.

If it's unsuitable, then do you know of any codes that I can use for ore slurry pipeline?

2) As for the S, I calculated it using the formula S = S x E x Sy, which was included in B31.4. The calculation gave the results of 43300 psi. However, using the value for S in the table shown in B31.3, the value is 25100 psi for pipe temperatures below 100 F. Can I use the latter value instead even though I'm using B31.4 and not B31.3 as code?

3) The slurry is indeed abrasive, but the velocity is as designed by the pipeline constructor. Thank you for the tip, I'll look for a way to determine the abrasiveness.

 

[1503-44]

Well, as GPT CHAT says, sorry for the confusion. I did not know the new B31,4 was revised to include slurry pipelines. Thanks for letting me know. Apparently you can use B31.4, if the owner agrees.

If you have a bad erosion problem, you may need the extra wall thickness anyway and the cost savings of using B31.4 wall thicknesses may be reduced considerably when you add a compensating thickness for corrosion/erosion allowance, or include any cost of internal coating. You have to do the economics study to get the answer. You also want to be sure that water hammer issues are minimal. They can be high with slurries.

B31.4, being a pipeline code that addresses any situations and requirements specifically for pipelines, I would think that would be advantagous in that regard as well.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[LittleInch]

Well the issues are many and various here:

7 years vs 15 eh? Sounds like they've replaced this line before now?

A slurry of 66% solids is extraordinary. Also what is in the slurry and its size and hardness all make it VERY difficult to predict life, especially as these may change gradually over time. The other issue is usually that you need to keep this goop travelling at a decent velocity in order to not settle out and clog, but that means excess wear.

Mr 44 - B 31.4 changed title to "liquids and slurries" in 2012....

This mysterious MAT is actually the design thickness for pressure containment in the way you're using it. This implies that they added a whole heap of allowance for erosion - How much?

Are you sure about that 4,500 psi for the slurry??? That's 310 bar, which is beyond even ASME class 2500 flanges so you're into well head flanges.
Really??

I agree with the 8.7mm for the B 31.4, but the S value in B31.3 comes from the table in B 31.3 for X60 pipe, which is 21500 psi. This one BIG reason you don't use B31.3 for pipeline high strength materials as it uses a MUCH lower value of stress in the calculation. DO NOT MIX AND MATCH DESIGN CODES. So no, you can't use one value of stress from one code in a different code.

The issue is more about what type of wall thinning you get. There is a lot of experience about pitting type corrosion, but I'm guessing your wall thinning is more even and more like a bigger area of wall thinning, especially at bends. This is quite unusual so using the min design thickness is probably as good a number as any other to decide if this section is close to failure.

Slurry pipeline design is a niche element and often needs rubber lined or PE lined pipe in order to survive more than a few years.

This presentation might help. I'll post a design manual as well.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[LittleInch]

A lot of the systems look at pressure drop and pump design, not erosion.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[jackieray]

@1503-44

Thanks for the advice. The issue I found here is that when I use B31.4 to calculate the MAT for 4500 psi, I get 0.344 in (8.73 mm) and when I use B31.3 I get 0.554 in (14.07 mm). It is understandable that calculations using B31.4 will give thinner MAT than B31.3 due to the difference in the S value of the 2 codes.

The pipeline engineers here ultimately use B31.3 to calculate MAT, however, and yet their MAT for 4500 psi is 0.270 in (6.87 mm) which is even thinner than my calculations using B31.4. I suspect that there are a big difference in S value between their calculations and mine.

I haven't had the chance to discuss thoroughly with them though, so it is just a hypothesis for now. Unfortunately, I don't know how much of an extra allowance for the pipe wall that the pipeline constructor used. The engineers here said that the extra protections are 1) coating and 2) cathodic protection.

 

[LittleInch]

Ha, Coating and CP are mitigations for EXTERNAL corrosion, not internal EROSION.

The only way anyone can get to a wt of 6.87 mm at 4,500 psi is to ignore the design factor in the B 31.4 equation ( i.e. basically going to yield stress) OR they are using a different pressure or someone is applying a factor based on the fact that the erosion isn't over the entire circumference. Or maybe they don't know what they are talking about. or maybe you've got hold of incomplete information. Who knows?

That 4,500 psi is unreal and needs confirmation as it looks wrong to me. What sort of profile are you pumping over?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

If you use 31.4 with 60,000 psi as the yield stress, the hoop stress wall thickness with DF=0.72 is 0.34"

Without their calculations, I can only guess that they used 60,000 as the allowable stress without the 0.4 factor. TOTALLY WRONG!

Since CP is a basic requirement, I wouldn't exactly call that "additional" protection.
You must find out what coating was used, then evaluate the effe tiveness of that. It doesn't sound like it worked very well.

LI yeah, ur right. Guess I haven't read the cover page since 2010.
CFR has not changed scope ever.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[jackieray]

@LittleInch

Thank you for your response. I will try to answer to the best of my knowledge as I had only started working here.

1) Yes. The company has been running for quite some time, so the pipelines (we have a total of 4 ore slurry pipelines here) have been completely replaced several times before.

2) Yes, the percent solids must be maintained at approx. 66% otherwise it would be less than economical to transport the slurry. Too little of it will increase the power needed in the dewatering process, and if it is well above 66% it will result in the increased difficulty in transporting the slurry and maybe the hypothetical wear rate of the pipeline due to the ore solids. As per the velocity I haven't calculated them yet but for the flowrate it is usually around 350-480 GPM. It can reach 500 at times, though.

3) "...This implies that they added a whole heap of allowance for erosion - How much?"

Sorry for my bad comprehension (English isn't my first language), but I don't quite get this question. I'll try answering it, though.

As far as I know, they don't have any special allowance for wear (abrasion, erosion, corrosion, etc), at least the pipeline engineers here didn't mention it to me. The measurements of the pipe they're using are somewhat uncommon, though. The inside diameter of the pipe they use in calculations is 0.438 in, even though for 6 NPS Sch 80 pipe normally it would be 0.432 in. I don't know if it's an allowance or not, but if it was, I'm not sure that extra 0.006 in of wall thickness is going to provide extra protection from erosion.

4) From what I saw in their calculations, in some sections of the pipelines, due to the steep hills, the pressure value used can get as high as 4100-4500 psi (they did say that the values came from the pipeline constructor though). 4500 psi is the maximum value, so I calculated the MAT using it as worst-case scenario. I am gonna discuss more with them regarding this ASAP.

However, I want to ask whether you can determine the Maximum Allowable Working Pressure (MAWP) for the ore slurry inside the pipelines by using the rearranged MAT formula, like this?

P = 2St/D

Or do I need to use other equations?

5) Thank you for your input! I was afraid that I had made a mistake in my calculations, but it is reassuring to know that other people can also get the same result haha. Also thank you for the tip. I guess I will be using B31.4 then since it is specified for slurry. I will also discuss this further along with the pressure values.

6) There are proofs and images of pitting type corrosion happening inside our ore slurry pipelines. However, I imagine that there will be a lot of wear that looks like scratches.

Can I use the minimum design thickness/MAT of 4500 psi for all sections of the pipe? So that I can propose that 0.344 in (8.73 mm) is the critical thickness that they have to maintain, before the pipelines fail?

7) Thank you. From what I read, it is favorable to use HDPE or HDPE-lined pipes in transporting slurries. However I don't think that the inside of the pipelines are lined with anything at all. I am gonna check into this.

Lastly, thank you so much for the references!!! I find that the references about slurry pipelines are quite scarce (or maybe I'm bad at researching lol) so it is a big help. Thank you once again!

 

[jackieray]

@LittleInch & @1503-44

Thanks for the swift responses!

1503-44, what is the 0.4 factor you mentioned? Is it the coefficient (Y) or design factor (E)?
The Y for ferritic steel under 482 F is 0.4 but I didn't use it since it is only specified in B31.3.
For the design factor, I did use 0.72 for E as B31.4 stated that it cannot exceed 0.72.

LittleInch, thanks for the reminder. I realize that I am still new and I haven't still fully understand the way the pipeline engineers here did the calculations. I will try to discuss with them more. I am also curious as to why they used B31.3 and still got thinner MAT than my calculations using B31.4.

I cannot post the calculations file or the screenshots here due to confidentiality issues, but here's an example of their calculation results.

Pipe Sections (Mile) | Maximum Pressure (psi) | MAT with 12.5% tolerance (mm)​

38 | 4500 | 6.870
37 | 4400 | 6.722
36 | 4300 | 6.573
etc.​

I am now starting to wonder if the Maximum Pressure is not the pressure readings of the slurry, but instead the Maximum Allowable Working Pressure(s) that are defined by the pipeline constructor.

However, why do the sections have differing MAWPs?

 

[GBTorpenhow]

However, why do the sections have differing MAWPs?

What is the elevation profile look like over the length of the pipeline? "Mountainous terrain" sounds like significant elevation differences.
A slurry of water and rock at 66% rock would be pretty dense, static pressure could change by 80-90 psi per 100 ft of elevation fairly easily, depending on the ore density. For the section shown 100 psi of increased operating pressure per mile is maybe 120 ft of elevation change per mile, guessing that mile 38 is lower elevation than mile 36 by something like that amount.

 

[jackieray]

@GBTorpenhow

Thank you for your response.

Yes, there are sections that are installed on steep hills. For Mile 38-37-36 though, from the elevation profile that I have, they have the same elevation level. May I ask what equation we can use to determine the ore slurry pressure based on % solids and elevation level?

However, those three and some the following sections are directly located AFTER said steep hills. The pipeline engineers said that in the sections on steep hills, the flow inside the pipe becomes a slack flow. I'm guessing that the high pressure may result from this said slack flow, and I'm thinking if a choke station installed at the end of the hills (just before Mile 38) can help alleviate the high pressure issue.

 

[GBTorpenhow]

May I ask what equation we can use to determine the ore slurry pressure based on % solids and elevation level?

Not the operating pressure but the difference in possible static pressure just due to the weight of the fluid column. This would be the difference in elevation (head) converted to pressure based on the density of the slurry. Static head isn't the only player here, though. If you don't have a good grasp on the extended Bernoulli equation I would start there.

The actual operating pressure of the pipeline is a completely different animal than the maximum design pressure. E.g. you reference slack flow, sections with slack flow would require operating pressures below the vapor pressure of the water component of the slurry, so there would be very low operating pressures in a section under this flow regime.

 

[LittleInch]

jackieray,

from your earlier post

1) I'm surprised therefore that they haven't invested in either rubber lined pipe or PE lined steel pipe. You clearly need steel as its very high pressure in places.
2) Generally speaking this isn't bad - a velocity of around 1.8m/sec max. What is the max particle size?
3) It looks to me like they didn't really add much for internal erosion, which is not a good move with slurry pipelines
4) The fact they can't easily tell you and are relying on anecdotal data from the pipeline constructor (not the designer?) tells me your pipeline engineers are really operators and that they don't properly understand pipeline design...
Yes you can use that formula so long as S is SMYS x DF.
5) Ok
6) Interesting
7) That's probably why they don't last very long

I can't answer the question about how they get to a lower MAT without seeing how they calculated it. However no pipeline Engineer worth his salt would use B 31.3 for this purpose.
Can you summarise the profile or give us a set of distance vs heights or at worst high points and low points? without we're going to struggle to help you.

Liquid line sin hilly areas often have different MAWP and sometimes different wall thicknesses. This is due to the static head varying a lot with the profile. If you've got valleys and peaks say 500m elevation difference, then the pressure at the bottom will be a lot higher than the highest point. It's not very common, but can happen in the right location.

You need to find the slurry density to work anything out

Slack flow on a slurry line is bad news as the local velocity can increase by a factor of 5 or more in the slack sections as the fluid free flows down from the high point. However slurry control valves are hard to find and make work because they wear away so fast.

Slurry lines are typically associated with mines and quarries and is a peculiar sub set of pipeline engineering, hence why there isn't much about and the companies tend to keep their experiences in house.

Give us a much data as you can and we can help you more.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

A slack line is extreme low pressure ( probably 15psia or less) at high points.
The fluid just makes it over the top of the hill. The pipe isn't even full at those points. It is similar to open channel flow and in fact the fluid flow is more like a waterfall, the pipe ahead is only partially full, so the water behind is free to accelerate to try to fill any empty space in the pipe ahead as it moves down the slope. There is no pipe full of water ahead to hold it back. On a steep slope this can increase velocity considerably, which quickly erodes the pipe in those locations.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[1503-44]

There is another possible source of wall thickness error.
How sure are you that the maximum pressure is 4500 psig.
It may help us to know the following.

Can you post the pipeline flow rate, diameter and elevation profile.
Can you post the density of the slurry, or at least the density of the ore.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[jackieray]

@GBTorpenhow

Thank you for the tip. Will look into Bernoulli eqs. As for the slack flow, it's something I've yet to confirm as it's (again) something that I heard from the engineers here. We do have pressure gauges and transmitters installed in several sections of the pipelines, so I'll look into it.

@LittleInch


1) Noted, I'll be looking into using liners as recommendations. However, some of our process pipes are lined with either rubber or ceramic.

2) What formula did you use to calculate the velocity? I used the simple V = flowrate (Q) / pipe cross sectional area (A) and got vastly different results. As for the max particle size, I don't have the exact data yet and I don't know if we have the means to determine it here, but generally 99% of the slurry pass 150-mesh screen.

3) Yes. I'll look into it.

4) Oh sorry, what I really meant as "constructor" is referring to the designer. As the pipelines were installed a long time ago, I don't know if the designers are still doing technical services work on them.

I also haven't looked at their calculations file yet as supposedly the data is confidential. Will be visiting the engineers soon though so hopefully I'll get hold of it. I'm gonna upload a rough sketch of the elevation profile later, I'm currently away from my PC.

Thanks with the info. In alignment with GBTorpenhow response, the engineers do use differing MAWPs for each mile of the pipe, hence the difference of MAT that I posted earlier.

I asked my direct superior here and the percent solids of the ore slurry is treated as "density" here. It's apparently a common practice in mineral processing plants to treat it as such.

For the slack flow, they do have a valve at the end of the slack flow sections, but I don't know if it wear quickly. I'll ask the engineers about it later this week.

I'm gonna try to get my hands on the data. In the meantime, thank you guys for replying and discussing this with me!!!

 

[jackieray]

@1503-44

Yes, that's exactly what the engineers here mentioned: on slack flow areas, the ore slurry isn't filling up the pipe all the way to the top, resulting in higher velocity. We don't have any means to monitor the velocity, only pressure, but the pressure isn't as low as 15 psia. I'll look for the more accurate pressure data.

As for the pressure of 4500 psig, I'm just using the highest MAWP that the engineers use. It applies only for Mile 38 though.

For the flow rate, it's around 350-480 GPM, anything lower will result in the slurry settling in the pipe, and anything higher will quicken the wear process of the pipe & affect the dewatering process too.

For the diameter of the pipe I have posted it above, along with the specifications of the pipe. I'm gonna upload the simple version of the elevation profile later though, as I'm away from my PC and it needs some retouching.

Sadly I don't think I can provide you guys with the density data. All we use here is percent solids for the slurry. I'll try asking if we have the equipments available to measure them, and maybe recommend that density analysis should be done regularly.

Thank you 44!