BIP148 is happening beginning on August 1st, 2017. But what does that mean for the average user?
For purposes of reading clarity, "legacy" means nodes not enforcing BIP148.
Risks for both legacy and BIP148 nodes (and the wallets trusting them)
If and only if BIP148 has minority hashrate support, there will be a chain split. Whether your node supports BIP148 or not, there is a risk that your economic counterparties (ie, people you want to pay and people you want to pay you) will be on the other side of the split. So long as nobody double-spends, this should be mostly okay for 100 blocks (about 16 hours); the only difference will be that transactions might confirm at different times. But if 100 blocks pass and miners begin spending their newly mined bitcoins (different on each side of the split), the chains' balances will begin to diverge, and transactions will become tied to one side or the other. There will be two "bitcoins".
Risks only for BIP148 nodes
Hey, there are none!
Risks only for legacy nodes
If the chain splits, then when / if / every time the BIP148 chain gets longer*, it will replace the legacy chain. Transactions that had confirmed on the legacy chain will become unconfirmed (unless they are also confirmed on the BIP148 chain). Bitcoins mined by legacy miners will cease to exist, as they lose their blocks. (This cannot occur in the inverse direction: no matter how long the legacy chain gets, BIP148 nodes will never let it reorg out the BIP148 chain.)
If the chain splits, the BIP148 side will have Segwit activating, whereas the legacy side will remain stagnated. There is a possibility this will give a market bias in favour of the BIP148 side, even independently from its initial adoption.
* Note that even a minority-hashrate chain can get longer than the majority-hashrate chain with some variance, although this becomes less probable with time. (On the other hand, the longer the chain split goes on, the more likely the BIP148 side grows in hashrate relative to the legacy side.)
Avoiding a chain split (and all the risks above)
A chain split can be avoided entirely if a sufficient amount of the economy adopts BIP148. Miners depend on their minted bitcoins in order to pay electric costs and recoup ASIC R&D costs. If the price they can sell their legacy bitcoins drops too significantly (possibly to zero when/if their blocks get reorg'd out by the BIP148 chain), they will have no choice but to switch to the BIP148 chain themselves, ensuring it is the longest chain for both BIP148 and legacy nodes. If the economy shows strong enough deployment of BIP148 prior to August 1st, it is even possible miners may switch preemptively, avoiding a chain split from occurring altogether. Note that only 51% of miners need to switch to the BIP148 chain to resolve or prevent the chain split, not the original 95% target.
BIP148 can also be automatically cancelled entirely by locking in Segwit before August 1st.
So there are a few ways a persistent chain split might be avoided:
95% of hashrate locks-in segwit before August 1st. No chain split at all.
51% of hashrate deploys BIP148 before August 1st. No chain split at all.
51% of hashrate deploys BIP148 after August 1st. Chain split gets resolved.
Legacy miners are compelled by the economy to switch to BIP148 after Aug 1st. Chain split gets resolved.
Basically, everyone’s risks go down with more people running BIP148 nodes.
source
https://np.reddit.com/r/Bitcoin/comments/6bxpsj/bip148_and_the_risks_it_entails_for_you_whether/ 
